net/ice/base: add flexible pipeline module
authorQi Zhang <qi.z.zhang@intel.com>
Wed, 16 Jan 2019 14:13:36 +0000 (22:13 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Fri, 18 Jan 2019 08:47:26 +0000 (09:47 +0100)
The flexible pipeline module provide the infrastructure for ice's
flexible packet processing feature.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
drivers/net/ice/Makefile
drivers/net/ice/base/ice_flex_pipe.c [new file with mode: 0644]
drivers/net/ice/base/ice_flex_pipe.h [new file with mode: 0644]
drivers/net/ice/base/ice_flex_type.h
drivers/net/ice/base/ice_type.h
drivers/net/ice/base/meson.build

index 4b421cd..1bb76ef 100644 (file)
@@ -49,6 +49,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_common.c
 SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_sched.c
 SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_switch.c
 SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_nvm.c
+SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_flex_pipe.c
 
 SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_ICE_PMD) += ice_rxtx.c
diff --git a/drivers/net/ice/base/ice_flex_pipe.c b/drivers/net/ice/base/ice_flex_pipe.c
new file mode 100644 (file)
index 0000000..1887fc3
--- /dev/null
@@ -0,0 +1,5018 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2019
+ */
+
+#include "ice_common.h"
+#include "ice_flex_pipe.h"
+#include "ice_protocol_type.h"
+#include "ice_flow.h"
+
+static const struct ice_tunnel_type_scan tnls[] = {
+       { TNL_VXLAN,            "TNL_VXLAN" },
+       { TNL_GTPC,             "TNL_GTPC" },
+       { TNL_GTPC_TEID,        "TNL_GTPC_TEID" },
+       { TNL_GTPU,             "TNL_GTPC" },
+       { TNL_GTPU_TEID,        "TNL_GTPU_TEID" },
+       { TNL_VXLAN_GPE,        "TNL_VXLAN_GPE" },
+       { TNL_GENEVE,           "TNL_GENEVE" },
+       { TNL_NAT,              "TNL_NAT" },
+       { TNL_ROCE_V2,          "TNL_ROCE_V2" },
+       { TNL_MPLSO_UDP,        "TNL_MPLSO_UDP" },
+       { TNL_UDP2_END,         "TNL_UDP2_END" },
+       { TNL_UPD_END,          "TNL_UPD_END" },
+       { TNL_LAST,             "" }
+};
+
+static const u32 ice_sect_lkup[ICE_BLK_COUNT][ICE_SECT_COUNT] = {
+       /* SWITCH */
+       {
+               ICE_SID_XLT0_SW,
+               ICE_SID_XLT_KEY_BUILDER_SW,
+               ICE_SID_XLT1_SW,
+               ICE_SID_XLT2_SW,
+               ICE_SID_PROFID_TCAM_SW,
+               ICE_SID_PROFID_REDIR_SW,
+               ICE_SID_FLD_VEC_SW,
+               ICE_SID_CDID_KEY_BUILDER_SW,
+               ICE_SID_CDID_REDIR_SW
+       },
+
+       /* ACL */
+       {
+               ICE_SID_XLT0_ACL,
+               ICE_SID_XLT_KEY_BUILDER_ACL,
+               ICE_SID_XLT1_ACL,
+               ICE_SID_XLT2_ACL,
+               ICE_SID_PROFID_TCAM_ACL,
+               ICE_SID_PROFID_REDIR_ACL,
+               ICE_SID_FLD_VEC_ACL,
+               ICE_SID_CDID_KEY_BUILDER_ACL,
+               ICE_SID_CDID_REDIR_ACL
+       },
+
+       /* FD */
+       {
+               ICE_SID_XLT0_FD,
+               ICE_SID_XLT_KEY_BUILDER_FD,
+               ICE_SID_XLT1_FD,
+               ICE_SID_XLT2_FD,
+               ICE_SID_PROFID_TCAM_FD,
+               ICE_SID_PROFID_REDIR_FD,
+               ICE_SID_FLD_VEC_FD,
+               ICE_SID_CDID_KEY_BUILDER_FD,
+               ICE_SID_CDID_REDIR_FD
+       },
+
+       /* RSS */
+       {
+               ICE_SID_XLT0_RSS,
+               ICE_SID_XLT_KEY_BUILDER_RSS,
+               ICE_SID_XLT1_RSS,
+               ICE_SID_XLT2_RSS,
+               ICE_SID_PROFID_TCAM_RSS,
+               ICE_SID_PROFID_REDIR_RSS,
+               ICE_SID_FLD_VEC_RSS,
+               ICE_SID_CDID_KEY_BUILDER_RSS,
+               ICE_SID_CDID_REDIR_RSS
+       },
+
+       /* PE */
+       {
+               ICE_SID_XLT0_PE,
+               ICE_SID_XLT_KEY_BUILDER_PE,
+               ICE_SID_XLT1_PE,
+               ICE_SID_XLT2_PE,
+               ICE_SID_PROFID_TCAM_PE,
+               ICE_SID_PROFID_REDIR_PE,
+               ICE_SID_FLD_VEC_PE,
+               ICE_SID_CDID_KEY_BUILDER_PE,
+               ICE_SID_CDID_REDIR_PE
+       }
+};
+
+/**
+ * ice_sect_id - returns section ID
+ * @blk: block type
+ * @sect: section type
+ *
+ * This helper function returns the proper section ID given a block type and a
+ * section type.
+ */
+static u32 ice_sect_id(enum ice_block blk, enum ice_sect sect)
+{
+       return ice_sect_lkup[blk][sect];
+}
+
+/**
+ * ice_pkg_val_buf
+ * @buf: pointer to the ice buffer
+ *
+ * This helper function validates a buffer's header.
+ */
+static struct ice_buf_hdr *ice_pkg_val_buf(struct ice_buf *buf)
+{
+       struct ice_buf_hdr *hdr;
+       u16 section_count;
+       u16 data_end;
+
+       hdr = (struct ice_buf_hdr *)buf->buf;
+       /* verify data */
+       section_count = LE16_TO_CPU(hdr->section_count);
+       if (section_count < ICE_MIN_S_COUNT || section_count > ICE_MAX_S_COUNT)
+               return NULL;
+
+       data_end = LE16_TO_CPU(hdr->data_end);
+       if (data_end < ICE_MIN_S_DATA_END || data_end > ICE_MAX_S_DATA_END)
+               return NULL;
+
+       return hdr;
+}
+
+/**
+ * ice_find_buf_table
+ * @ice_seg: pointer to the ice segment
+ *
+ * Returns the address of the buffer table within the ice segment.
+ */
+static struct ice_buf_table *ice_find_buf_table(struct ice_seg *ice_seg)
+{
+       struct ice_nvm_table *nvms;
+
+       nvms = (struct ice_nvm_table *)(ice_seg->device_table +
+               LE32_TO_CPU(ice_seg->device_table_count));
+
+       return (struct ice_buf_table *)
+               (nvms->vers + LE32_TO_CPU(nvms->table_count));
+}
+
+/**
+ * ice_pkg_enum_buf
+ * @ice_seg: pointer to the ice segment (or NULL on subsequent calls)
+ * @state: pointer to the enum state
+ *
+ * This function will enumerate all the buffers in the ice segment. The first
+ * call is made with the ice_seg parameter non-NULL; on subsequent calls,
+ * ice_seg is set to NULL which continues the enumeration. When the function
+ * returns a NULL pointer, then the end of the buffers has been reached, or an
+ * unexpected value has been detected (for example an invalid section count or
+ * an invalid buffer end value).
+ */
+static struct ice_buf_hdr *
+ice_pkg_enum_buf(struct ice_seg *ice_seg, struct ice_pkg_enum *state)
+{
+       if (ice_seg) {
+               state->buf_table = ice_find_buf_table(ice_seg);
+               if (!state->buf_table)
+                       return NULL;
+
+               state->buf_idx = 0;
+               return ice_pkg_val_buf(state->buf_table->buf_array);
+       }
+
+       if (++state->buf_idx < LE32_TO_CPU(state->buf_table->buf_count))
+               return ice_pkg_val_buf(state->buf_table->buf_array +
+                                      state->buf_idx);
+       else
+               return NULL;
+}
+
+/**
+ * ice_pkg_advance_sect
+ * @ice_seg: pointer to the ice segment (or NULL on subsequent calls)
+ * @state: pointer to the enum state
+ *
+ * This helper function will advance the section within the ice segment,
+ * also advancing the buffer if needed.
+ */
+static bool
+ice_pkg_advance_sect(struct ice_seg *ice_seg, struct ice_pkg_enum *state)
+{
+       if (!ice_seg && !state->buf)
+               return false;
+
+       if (!ice_seg && state->buf)
+               if (++state->sect_idx < LE16_TO_CPU(state->buf->section_count))
+                       return true;
+
+       state->buf = ice_pkg_enum_buf(ice_seg, state);
+       if (!state->buf)
+               return false;
+
+       /* start of new buffer, reset section index */
+       state->sect_idx = 0;
+       return true;
+}
+
+/**
+ * ice_pkg_enum_section
+ * @ice_seg: pointer to the ice segment (or NULL on subsequent calls)
+ * @state: pointer to the enum state
+ * @sect_type: section type to enumerate
+ *
+ * This function will enumerate all the sections of a particular type in the
+ * ice segment. The first call is made with the ice_seg parameter non-NULL;
+ * on subsequent calls, ice_seg is set to NULL which continues the enumeration.
+ * When the function returns a NULL pointer, then the end of the matching
+ * sections has been reached.
+ */
+static void *
+ice_pkg_enum_section(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
+                    u32 sect_type)
+{
+       u16 offset, size;
+
+       if (ice_seg)
+               state->type = sect_type;
+
+       if (!ice_pkg_advance_sect(ice_seg, state))
+               return NULL;
+
+       /* scan for next matching section */
+       while (state->buf->section_entry[state->sect_idx].type !=
+              CPU_TO_LE32(state->type))
+               if (!ice_pkg_advance_sect(NULL, state))
+                       return NULL;
+
+       /* validate section */
+       offset = LE16_TO_CPU(state->buf->section_entry[state->sect_idx].offset);
+       if (offset < ICE_MIN_S_OFF || offset > ICE_MAX_S_OFF)
+               return NULL;
+
+       size = LE16_TO_CPU(state->buf->section_entry[state->sect_idx].size);
+       if (size < ICE_MIN_S_SZ || size > ICE_MAX_S_SZ)
+               return NULL;
+
+       /* make sure the section fits in the buffer */
+       if (offset + size > ICE_PKG_BUF_SIZE)
+               return NULL;
+
+       state->sect_type =
+               LE32_TO_CPU(state->buf->section_entry[state->sect_idx].type);
+
+       /* calc pointer to this section */
+       state->sect = ((u8 *)state->buf) +
+               LE16_TO_CPU(state->buf->section_entry[state->sect_idx].offset);
+
+       return state->sect;
+}
+
+/**
+ * ice_pkg_enum_entry
+ * @ice_seg: pointer to the ice segment (or NULL on subsequent calls)
+ * @state: pointer to the enum state
+ * @sect_type: section type to enumerate
+ * @offset: pointer to variable that receives the offset in the table (optional)
+ * @handler: function that handles access to the entries into the section type
+ *
+ * This function will enumerate all the entries in particular section type in
+ * the ice segment. The first call is made with the ice_seg parameter non-NULL;
+ * on subsequent calls, ice_seg is set to NULL which continues the enumeration.
+ * When the function returns a NULL pointer, then the end of the entries has
+ * been reached.
+ *
+ * Since each section may have a different header and entry size, the handler
+ * function is needed to determine the number and location entries in each
+ * section.
+ *
+ * The offset parameter is optional, but should be used for sections that
+ * contain an offset for each section table. For such cases, the section handler
+ * function must return the appropriate offset + index to give the absolution
+ * offset for each entry. For example, if the base for a section's header
+ * indicates a base offset of 10, and the index for the entry is 2, then
+ * section handler function should set the offset to 10 + 2 = 12.
+ */
+static void *
+ice_pkg_enum_entry(struct ice_seg *ice_seg, struct ice_pkg_enum *state,
+                  u32 sect_type, u32 *offset,
+                  void *(*handler)(u32 sect_type, void *section,
+                                   u32 index, u32 *offset))
+{
+       void *entry;
+
+       if (ice_seg) {
+               if (!handler)
+                       return NULL;
+
+               if (!ice_pkg_enum_section(ice_seg, state, sect_type))
+                       return NULL;
+
+               state->entry_idx = 0;
+               state->handler = handler;
+       } else {
+               state->entry_idx++;
+       }
+
+       if (!state->handler)
+               return NULL;
+
+       /* get entry */
+       entry = state->handler(state->sect_type, state->sect, state->entry_idx,
+                              offset);
+       if (!entry) {
+               /* end of a section, look for another section of this type */
+               if (!ice_pkg_enum_section(NULL, state, 0))
+                       return NULL;
+
+               state->entry_idx = 0;
+               entry = state->handler(state->sect_type, state->sect,
+                                      state->entry_idx, offset);
+       }
+
+       return entry;
+}
+
+/**
+ * ice_boost_tcam_handler
+ * @sect_type: section type
+ * @section: pointer to section
+ * @index: index of the boost tcam entry to be returned
+ * @offset: pointer to receive absolute offset, always 0 for boost tcam sections
+ *
+ * This is a callback function that can be passed to ice_pkg_enum_entry.
+ * Handles enumeration of individual boost tcam entries.
+ */
+static void *
+ice_boost_tcam_handler(u32 sect_type, void *section, u32 index, u32 *offset)
+{
+       struct ice_boost_tcam_section *boost;
+
+       if (!section)
+               return NULL;
+
+       if (sect_type != ICE_SID_RXPARSER_BOOST_TCAM)
+               return NULL;
+
+       if (index > ICE_MAX_BST_TCAMS_IN_BUF)
+               return NULL;
+
+       if (offset)
+               *offset = 0;
+
+       boost = (struct ice_boost_tcam_section *)section;
+       if (index >= LE16_TO_CPU(boost->count))
+               return NULL;
+
+       return boost->tcam + index;
+}
+
+/**
+ * ice_find_boost_entry
+ * @ice_seg: pointer to the ice segment (non-NULL)
+ * @addr: Boost TCAM address of entry to search for
+ * @entry: returns pointer to the entry
+ *
+ * Finds a particular Boost TCAM entry and returns a pointer to that entry
+ * if it is found. The ice_seg parameter must not be NULL since the first call
+ * to ice_pkg_enum_entry requires a pointer to an actual ice_segment structure.
+ */
+static enum ice_status
+ice_find_boost_entry(struct ice_seg *ice_seg, u16 addr,
+                    struct ice_boost_tcam_entry **entry)
+{
+       struct ice_boost_tcam_entry *tcam;
+       struct ice_pkg_enum state;
+
+       ice_memset(&state, 0, sizeof(state), ICE_NONDMA_MEM);
+
+       if (!ice_seg)
+               return ICE_ERR_PARAM;
+
+       do {
+               tcam = (struct ice_boost_tcam_entry *)
+                      ice_pkg_enum_entry(ice_seg, &state,
+                                         ICE_SID_RXPARSER_BOOST_TCAM, NULL,
+                                         ice_boost_tcam_handler);
+               if (tcam && LE16_TO_CPU(tcam->addr) == addr) {
+                       *entry = tcam;
+                       return ICE_SUCCESS;
+               }
+
+               ice_seg = NULL;
+       } while (tcam);
+
+       *entry = NULL;
+       return ICE_ERR_CFG;
+}
+
+/**
+ * ice_label_enum_handler
+ * @sect_type: section type
+ * @section: pointer to section
+ * @index: index of the label entry to be returned
+ * @offset: pointer to receive absolute offset, always zero for label sections
+ *
+ * This is a callback function that can be passed to ice_pkg_enum_entry.
+ * Handles enumeration of individual label entries.
+ */
+static void *
+ice_label_enum_handler(u32 __always_unused sect_type, void *section, u32 index,
+                      u32 *offset)
+{
+       struct ice_label_section *labels;
+
+       if (!section)
+               return NULL;
+
+       if (index > ICE_MAX_LABELS_IN_BUF)
+               return NULL;
+
+       if (offset)
+               *offset = 0;
+
+       labels = (struct ice_label_section *)section;
+       if (index >= LE16_TO_CPU(labels->count))
+               return NULL;
+
+       return labels->label + index;
+}
+
+/**
+ * ice_enum_labels
+ * @ice_seg: pointer to the ice segment (NULL on subsequent calls)
+ * @type: the section type that will contain the label (0 on subsequent calls)
+ * @state: ice_pkg_enum structure that will hold the state of the enumeration
+ * @value: pointer to a value that will return the label's value if found
+ *
+ * Enumerates a list of labels in the package. The caller will call
+ * ice_enum_labels(ice_seg, type, ...) to start the enumeration, then call
+ * ice_enum_labels(NULL, 0, ...) to continue. When the function returns a NULL
+ * the end of the list has been reached.
+ */
+static char *
+ice_enum_labels(struct ice_seg *ice_seg, u32 type, struct ice_pkg_enum *state,
+               u16 *value)
+{
+       struct ice_label *label;
+
+       /* Check for valid label section on first call */
+       if (type && !(type >= ICE_SID_LBL_FIRST && type <= ICE_SID_LBL_LAST))
+               return NULL;
+
+       label = (struct ice_label *)ice_pkg_enum_entry(ice_seg, state, type,
+                                                      NULL,
+                                                      ice_label_enum_handler);
+       if (!label)
+               return NULL;
+
+       *value = LE16_TO_CPU(label->value);
+       return label->name;
+}
+
+/**
+ * ice_init_pkg_hints
+ * @hw: pointer to the HW structure
+ * @ice_seg: pointer to the segment of the package scan (non-NULL)
+ *
+ * This function will scan the package and save off relevant information
+ * (hints or metadata) for driver use. The ice_seg parameter must not be NULL
+ * since the first call to ice_enum_labels requires a pointer to an actual
+ * ice_seg structure.
+ */
+void ice_init_pkg_hints(struct ice_hw *hw, struct ice_seg *ice_seg)
+{
+       struct ice_pkg_enum state;
+       char *label_name;
+       u16 val;
+       int i;
+
+       ice_memset(&hw->tnl, 0, sizeof(hw->tnl), ICE_NONDMA_MEM);
+
+       if (!ice_seg)
+               return;
+
+       label_name = ice_enum_labels(ice_seg, ICE_SID_LBL_RXPARSER_TMEM, &state,
+                                    &val);
+
+       while (label_name && hw->tnl.count < ICE_TUNNEL_MAX_ENTRIES) {
+               for (i = 0; tnls[i].type != TNL_LAST; i++) {
+                       if (!strncmp(label_name, tnls[i].label_prefix,
+                                    strlen(tnls[i].label_prefix))) {
+                               hw->tnl.tbl[hw->tnl.count].type = tnls[i].type;
+                               hw->tnl.tbl[hw->tnl.count].valid = false;
+                               hw->tnl.tbl[hw->tnl.count].in_use = false;
+                               hw->tnl.tbl[hw->tnl.count].marked = false;
+                               hw->tnl.tbl[hw->tnl.count].boost_addr = val;
+                               hw->tnl.tbl[hw->tnl.count].port = 0;
+                               hw->tnl.count++;
+                               break;
+                       }
+               }
+
+               label_name = ice_enum_labels(NULL, 0, &state, &val);
+       }
+
+       /* Cache the appropriate boost tcam entry pointers */
+       for (i = 0; i < hw->tnl.count; i++) {
+               ice_find_boost_entry(ice_seg, hw->tnl.tbl[i].boost_addr,
+                                    &hw->tnl.tbl[i].boost_entry);
+               if (hw->tnl.tbl[i].boost_entry)
+                       hw->tnl.tbl[i].valid = true;
+       }
+}
+
+/* Key creation */
+
+#define ICE_DC_KEY     0x1     /* don't care */
+#define ICE_DC_KEYINV  0x1
+#define ICE_NM_KEY     0x0     /* never match */
+#define ICE_NM_KEYINV  0x0
+#define ICE_0_KEY      0x1     /* match 0 */
+#define ICE_0_KEYINV   0x0
+#define ICE_1_KEY      0x0     /* match 1 */
+#define ICE_1_KEYINV   0x1
+
+/**
+ * ice_gen_key_word - generate 16-bits of a key/mask word
+ * @val: the value
+ * @valid: valid bits mask (change only the valid bits)
+ * @dont_care: don't care mask
+ * @nvr_mtch: never match mask
+ * @key: pointer to an array of where the resulting key portion
+ * @key_inv: pointer to an array of where the resulting key invert portion
+ *
+ * This function generates 16-bits from a 8-bit value, an 8-bit don't care mask
+ * and an 8-bit never match mask. The 16-bits of output are divided into 8 bits
+ * of key and 8 bits of key invert.
+ *
+ *     '0' =    b01, always match a 0 bit
+ *     '1' =    b10, always match a 1 bit
+ *     '?' =    b11, don't care bit (always matches)
+ *     '~' =    b00, never match bit
+ *
+ * Input:
+ *          val:         b0  1  0  1  0  1
+ *          dont_care:   b0  0  1  1  0  0
+ *          never_mtch:  b0  0  0  0  1  1
+ *          ------------------------------
+ * Result:  key:        b01 10 11 11 00 00
+ */
+static enum ice_status
+ice_gen_key_word(u8 val, u8 valid, u8 dont_care, u8 nvr_mtch, u8 *key,
+                u8 *key_inv)
+{
+       u8 in_key = *key, in_key_inv = *key_inv;
+       u8 i;
+
+       /* 'dont_care' and 'nvr_mtch' masks cannot overlap */
+       if ((dont_care ^ nvr_mtch) != (dont_care | nvr_mtch))
+               return ICE_ERR_CFG;
+
+       *key = 0;
+       *key_inv = 0;
+
+       /* encode the 8 bits into 8-bit key and 8-bit key invert */
+       for (i = 0; i < 8; i++) {
+               *key >>= 1;
+               *key_inv >>= 1;
+
+               if (!(valid & 0x1)) { /* change only valid bits */
+                       *key |= (in_key & 0x1) << 7;
+                       *key_inv |= (in_key_inv & 0x1) << 7;
+               } else if (dont_care & 0x1) { /* don't care bit */
+                       *key |= ICE_DC_KEY << 7;
+                       *key_inv |= ICE_DC_KEYINV << 7;
+               } else if (nvr_mtch & 0x1) { /* never match bit */
+                       *key |= ICE_NM_KEY << 7;
+                       *key_inv |= ICE_NM_KEYINV << 7;
+               } else if (val & 0x01) { /* exact 1 match */
+                       *key |= ICE_1_KEY << 7;
+                       *key_inv |= ICE_1_KEYINV << 7;
+               } else { /* exact 0 match */
+                       *key |= ICE_0_KEY << 7;
+                       *key_inv |= ICE_0_KEYINV << 7;
+               }
+
+               dont_care >>= 1;
+               nvr_mtch >>= 1;
+               valid >>= 1;
+               val >>= 1;
+               in_key >>= 1;
+               in_key_inv >>= 1;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_bits_max_set - determine if the number of bits set is within a maximum
+ * @mask: pointer to the byte array which is the mask
+ * @size: the number of bytes in the mask
+ * @max: the max number of set bits
+ *
+ * This function determines if there are at most 'max' number of bits set in an
+ * array. Returns true if the number for bits set is <= max or will return false
+ * otherwise.
+ */
+static bool ice_bits_max_set(const u8 *mask, u16 size, u16 max)
+{
+       u16 count = 0;
+       u16 i, j;
+
+       /* check each byte */
+       for (i = 0; i < size; i++) {
+               /* if 0, go to next byte */
+               if (!mask[i])
+                       continue;
+
+               /* We know there is at least one set bit in this byte because of
+                * the above check; if we already have found 'max' number of
+                * bits set, then we can return failure now.
+                */
+               if (count == max)
+                       return false;
+
+               /* count the bits in this byte, checking threshold */
+               for (j = 0; j < BITS_PER_BYTE; j++) {
+                       count += (mask[i] & (0x1 << j)) ? 1 : 0;
+                       if (count > max)
+                               return false;
+               }
+       }
+
+       return true;
+}
+
+/**
+ * ice_set_key - generate a variable sized key with multiples of 16-bits
+ * @key: pointer to where the key will be stored
+ * @size: the size of the complete key in bytes (must be even)
+ * @val: array of 8-bit values that makes up the value portion of the key
+ * @upd: array of 8-bit masks that determine what key portion to update
+ * @dc: array of 8-bit masks that make up the dont' care mask
+ * @nm: array of 8-bit masks that make up the never match mask
+ * @off: the offset of the first byte in the key to update
+ * @len: the number of bytes in the key update
+ *
+ * This function generates a key from a value, a don't care mask and a never
+ * match mask.
+ * upd, dc, and nm are optional parameters, and can be NULL:
+ *     upd == NULL --> udp mask is all 1's (update all bits)
+ *     dc == NULL --> dc mask is all 0's (no don't care bits)
+ *     nm == NULL --> nm mask is all 0's (no never match bits)
+ */
+enum ice_status
+ice_set_key(u8 *key, u16 size, u8 *val, u8 *upd, u8 *dc, u8 *nm, u16 off,
+           u16 len)
+{
+       u16 half_size;
+       u16 i;
+
+       /* size must be a multiple of 2 bytes. */
+       if (size % 2)
+               return ICE_ERR_CFG;
+       half_size = size / 2;
+
+       if (off + len > half_size)
+               return ICE_ERR_CFG;
+
+       /* Make sure at most one bit is set in the never match mask. Having more
+        * than one never match mask bit set will cause HW to consume excessive
+        * power otherwise; this is a power management efficiency check.
+        */
+#define ICE_NVR_MTCH_BITS_MAX  1
+       if (nm && !ice_bits_max_set(nm, len, ICE_NVR_MTCH_BITS_MAX))
+               return ICE_ERR_CFG;
+
+       for (i = 0; i < len; i++)
+               if (ice_gen_key_word(val[i], upd ? upd[i] : 0xff,
+                                    dc ? dc[i] : 0, nm ? nm[i] : 0,
+                                    key + off + i, key + half_size + off + i))
+                       return ICE_ERR_CFG;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_acquire_global_cfg_lock
+ * @hw: pointer to the HW structure
+ * @access: access type (read or write)
+ *
+ * This function will request ownership of the global config lock for reading
+ * or writing of the package. When attempting to obtain write access, the
+ * caller must check for the following two return values:
+ *
+ * ICE_SUCCESS        - Means the caller has acquired the global config lock
+ *                      and can perform writing of the package.
+ * ICE_ERR_AQ_NO_WORK - Indicates another driver has already written the
+ *                      package or has found that no update was necessary; in
+ *                      this case, the caller can just skip performing any
+ *                      update of the package.
+ */
+static enum ice_status
+ice_acquire_global_cfg_lock(struct ice_hw *hw,
+                           enum ice_aq_res_access_type access)
+{
+       enum ice_status status;
+
+       ice_debug(hw, ICE_DBG_TRACE, "ice_acquire_global_cfg_lock");
+
+       status = ice_acquire_res(hw, ICE_GLOBAL_CFG_LOCK_RES_ID, access,
+                                ICE_GLOBAL_CFG_LOCK_TIMEOUT);
+
+       if (status == ICE_ERR_AQ_NO_WORK)
+               ice_debug(hw, ICE_DBG_PKG,
+                         "Global config lock: No work to do\n");
+
+       return status;
+}
+
+/**
+ * ice_release_global_cfg_lock
+ * @hw: pointer to the HW structure
+ *
+ * This function will release the global config lock.
+ */
+static void ice_release_global_cfg_lock(struct ice_hw *hw)
+{
+       ice_release_res(hw, ICE_GLOBAL_CFG_LOCK_RES_ID);
+}
+
+/**
+ * ice_acquire_change_lock
+ * @hw: pointer to the HW structure
+ * @access: access type (read or write)
+ *
+ * This function will request ownership of the change lock.
+ */
+static enum ice_status
+ice_acquire_change_lock(struct ice_hw *hw, enum ice_aq_res_access_type access)
+{
+       ice_debug(hw, ICE_DBG_TRACE, "ice_acquire_change_lock");
+
+       return ice_acquire_res(hw, ICE_CHANGE_LOCK_RES_ID, access,
+                              ICE_CHANGE_LOCK_TIMEOUT);
+}
+
+/**
+ * ice_release_change_lock
+ * @hw: pointer to the HW structure
+ *
+ * This function will release the change lock using the proper Admin Command.
+ */
+static void ice_release_change_lock(struct ice_hw *hw)
+{
+       ice_debug(hw, ICE_DBG_TRACE, "ice_release_change_lock");
+
+       ice_release_res(hw, ICE_CHANGE_LOCK_RES_ID);
+}
+
+/**
+ * ice_aq_download_pkg
+ * @hw: pointer to the hardware structure
+ * @pkg_buf: the package buffer to transfer
+ * @buf_size: the size of the package buffer
+ * @last_buf: last buffer indicator
+ * @error_offset: returns error offset
+ * @error_info: returns error information
+ * @cd: pointer to command details structure or NULL
+ *
+ * Download Package (0x0C40)
+ */
+static enum ice_status
+ice_aq_download_pkg(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf,
+                   u16 buf_size, bool last_buf, u32 *error_offset,
+                   u32 *error_info, struct ice_sq_cd *cd)
+{
+       struct ice_aqc_download_pkg *cmd;
+       struct ice_aq_desc desc;
+       enum ice_status status;
+
+       ice_debug(hw, ICE_DBG_TRACE, "ice_aq_download_pkg");
+
+       if (error_offset)
+               *error_offset = 0;
+       if (error_info)
+               *error_info = 0;
+
+       cmd = &desc.params.download_pkg;
+       ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_download_pkg);
+       desc.flags |= CPU_TO_LE16(ICE_AQ_FLAG_RD);
+
+       if (last_buf)
+               cmd->flags |= ICE_AQC_DOWNLOAD_PKG_LAST_BUF;
+
+       status = ice_aq_send_cmd(hw, &desc, pkg_buf, buf_size, cd);
+       if (status == ICE_ERR_AQ_ERROR) {
+               /* Read error from buffer only when the FW returned an error */
+               struct ice_aqc_download_pkg_resp *resp;
+
+               resp = (struct ice_aqc_download_pkg_resp *)pkg_buf;
+               if (error_offset)
+                       *error_offset = LE32_TO_CPU(resp->error_offset);
+               if (error_info)
+                       *error_info = LE32_TO_CPU(resp->error_info);
+       }
+
+       return status;
+}
+
+/**
+ * ice_aq_upload_section
+ * @hw: pointer to the hardware structure
+ * @pkg_buf: the package buffer which will receive the section
+ * @buf_size: the size of the package buffer
+ * @cd: pointer to command details structure or NULL
+ *
+ * Upload Section (0x0C41)
+ */
+enum ice_status
+ice_aq_upload_section(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf,
+                     u16 buf_size, struct ice_sq_cd *cd)
+{
+       struct ice_aq_desc desc;
+
+       ice_debug(hw, ICE_DBG_TRACE, "ice_aq_upload_section");
+       ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_upload_section);
+       desc.flags |= CPU_TO_LE16(ICE_AQ_FLAG_RD);
+
+       return ice_aq_send_cmd(hw, &desc, pkg_buf, buf_size, cd);
+}
+
+/**
+ * ice_aq_update_pkg
+ * @hw: pointer to the hardware structure
+ * @pkg_buf: the package cmd buffer
+ * @buf_size: the size of the package cmd buffer
+ * @last_buf: last buffer indicator
+ * @error_offset: returns error offset
+ * @error_info: returns error information
+ * @cd: pointer to command details structure or NULL
+ *
+ * Update Package (0x0C42)
+ */
+static enum ice_status
+ice_aq_update_pkg(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf, u16 buf_size,
+                 bool last_buf, u32 *error_offset, u32 *error_info,
+                 struct ice_sq_cd *cd)
+{
+       struct ice_aqc_download_pkg *cmd;
+       struct ice_aq_desc desc;
+       enum ice_status status;
+
+       ice_debug(hw, ICE_DBG_TRACE, "ice_aq_update_pkg");
+
+       if (error_offset)
+               *error_offset = 0;
+       if (error_info)
+               *error_info = 0;
+
+       cmd = &desc.params.download_pkg;
+       ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_update_pkg);
+       desc.flags |= CPU_TO_LE16(ICE_AQ_FLAG_RD);
+
+       if (last_buf)
+               cmd->flags |= ICE_AQC_DOWNLOAD_PKG_LAST_BUF;
+
+       status = ice_aq_send_cmd(hw, &desc, pkg_buf, buf_size, cd);
+       if (status == ICE_ERR_AQ_ERROR) {
+               /* Read error from buffer only when the FW returned an error */
+               struct ice_aqc_download_pkg_resp *resp;
+
+               resp = (struct ice_aqc_download_pkg_resp *)pkg_buf;
+               if (error_offset)
+                       *error_offset = LE32_TO_CPU(resp->error_offset);
+               if (error_info)
+                       *error_info = LE32_TO_CPU(resp->error_info);
+       }
+
+       return status;
+}
+
+/**
+ * ice_find_seg_in_pkg
+ * @hw: pointer to the hardware structure
+ * @seg_type: the segment type to search for (i.e., SEGMENT_TYPE_CPK)
+ * @pkg_hdr: pointer to the package header to be searched
+ *
+ * This function searches a package file for a particular segment type. On
+ * success it returns a pointer to the segment header, otherwise it will
+ * return NULL.
+ */
+struct ice_generic_seg_hdr *
+ice_find_seg_in_pkg(struct ice_hw *hw, u32 seg_type,
+                   struct ice_pkg_hdr *pkg_hdr)
+{
+       u32 i;
+
+       ice_debug(hw, ICE_DBG_TRACE, "%s\n", __func__);
+       ice_debug(hw, ICE_DBG_PKG, "Package version: %d.%d.%d.%d\n",
+                 pkg_hdr->format_ver.major, pkg_hdr->format_ver.minor,
+                 pkg_hdr->format_ver.update, pkg_hdr->format_ver.draft);
+
+       /* Search all package segments for the requested segment type */
+       for (i = 0; i < LE32_TO_CPU(pkg_hdr->seg_count); i++) {
+               struct ice_generic_seg_hdr *seg;
+
+               seg = (struct ice_generic_seg_hdr *)
+                       ((u8 *)pkg_hdr + LE32_TO_CPU(pkg_hdr->seg_offset[i]));
+
+               if (LE32_TO_CPU(seg->seg_type) == seg_type)
+                       return seg;
+       }
+
+       return NULL;
+}
+
+/**
+ * ice_update_pkg
+ * @hw: pointer to the hardware structure
+ * @bufs: pointer to an array of buffers
+ * @count: the number of buffers in the array
+ *
+ * Obtains change lock and updates package.
+ */
+enum ice_status
+ice_update_pkg(struct ice_hw *hw, struct ice_buf *bufs, u32 count)
+{
+       enum ice_status status;
+       u32 offset, info, i;
+
+       status = ice_acquire_change_lock(hw, ICE_RES_WRITE);
+       if (status)
+               return status;
+
+       for (i = 0; i < count; i++) {
+               bool last = ((i + 1) == count);
+
+               struct ice_buf_hdr *bh = (struct ice_buf_hdr *)(bufs + i);
+
+               status = ice_aq_update_pkg(hw, bh, LE16_TO_CPU(bh->data_end),
+                                          last, &offset, &info, NULL);
+
+               if (status) {
+                       ice_debug(hw, ICE_DBG_PKG,
+                                 "Update pkg failed: err %d off %d inf %d\n",
+                                 status, offset, info);
+                       break;
+               }
+       }
+
+       ice_release_change_lock(hw);
+
+       return status;
+}
+
+/**
+ * ice_dwnld_cfg_bufs
+ * @hw: pointer to the hardware structure
+ * @bufs: pointer to an array of buffers
+ * @count: the number of buffers in the array
+ *
+ * Obtains global config lock and downloads the package configuration buffers
+ * to the firmware. Metadata buffers are skipped, and the first metadata buffer
+ * found indicates that the rest of the buffers are all metadata buffers.
+ */
+static enum ice_status
+ice_dwnld_cfg_bufs(struct ice_hw *hw, struct ice_buf *bufs, u32 count)
+{
+       enum ice_status status;
+       struct ice_buf_hdr *bh;
+       u32 offset, info, i;
+
+       if (!bufs || !count)
+               return ICE_ERR_PARAM;
+
+       /* If the first buffer's first section has its metadata bit set
+        * then there are no buffers to be downloaded, and the operation is
+        * considered a success.
+        */
+       bh = (struct ice_buf_hdr *)bufs;
+       if (LE32_TO_CPU(bh->section_entry[0].type) & ICE_METADATA_BUF)
+               return ICE_SUCCESS;
+
+       status = ice_acquire_global_cfg_lock(hw, ICE_RES_WRITE);
+       if (status)
+               return status;
+
+       for (i = 0; i < count; i++) {
+               bool last = ((i + 1) == count);
+
+               if (!last) {
+                       /* check next buffer for metadata flag */
+                       bh = (struct ice_buf_hdr *)(bufs + i + 1);
+
+                       /* A set metadata flag in the next buffer will signal
+                        * that the current buffer will be the last buffer
+                        * downloaded
+                        */
+                       if (LE16_TO_CPU(bh->section_count))
+                               if (LE32_TO_CPU(bh->section_entry[0].type) &
+                                   ICE_METADATA_BUF)
+                                       last = true;
+               }
+
+               bh = (struct ice_buf_hdr *)(bufs + i);
+
+               status = ice_aq_download_pkg(hw, bh, LE16_TO_CPU(bh->data_end),
+                                            last, &offset, &info, NULL);
+
+               if (status) {
+                       ice_debug(hw, ICE_DBG_PKG,
+                                 "Pkg download failed: err %d off %d inf %d\n",
+                                 status, offset, info);
+                       break;
+               }
+
+               if (last)
+                       break;
+       }
+
+       ice_release_global_cfg_lock(hw);
+
+       return status;
+}
+
+/**
+ * ice_aq_get_pkg_info_list
+ * @hw: pointer to the hardware structure
+ * @pkg_info: the buffer which will receive the information list
+ * @buf_size: the size of the pkg_info information buffer
+ * @cd: pointer to command details structure or NULL
+ *
+ * Get Package Info List (0x0C43)
+ */
+static enum ice_status
+ice_aq_get_pkg_info_list(struct ice_hw *hw,
+                        struct ice_aqc_get_pkg_info_resp *pkg_info,
+                        u16 buf_size, struct ice_sq_cd *cd)
+{
+       struct ice_aq_desc desc;
+
+       ice_debug(hw, ICE_DBG_TRACE, "ice_aq_get_pkg_info_list");
+       ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_get_pkg_info_list);
+
+       return ice_aq_send_cmd(hw, &desc, pkg_info, buf_size, cd);
+}
+
+/**
+ * ice_download_pkg
+ * @hw: pointer to the hardware structure
+ * @ice_seg: pointer to the segment of the package to be downloaded
+ *
+ * Handles the download of a complete package.
+ */
+enum ice_status ice_download_pkg(struct ice_hw *hw, struct ice_seg *ice_seg)
+{
+       struct ice_buf_table *ice_buf_tbl;
+
+       ice_debug(hw, ICE_DBG_TRACE, "%s\n", __func__);
+       ice_debug(hw, ICE_DBG_PKG, "Segment version: %d.%d.%d.%d\n",
+                 ice_seg->hdr.seg_ver.major, ice_seg->hdr.seg_ver.minor,
+                 ice_seg->hdr.seg_ver.update, ice_seg->hdr.seg_ver.draft);
+
+       ice_debug(hw, ICE_DBG_PKG, "Seg: type 0x%X, size %d, name %s\n",
+                 LE32_TO_CPU(ice_seg->hdr.seg_type),
+                 LE32_TO_CPU(ice_seg->hdr.seg_size), ice_seg->hdr.seg_name);
+
+       ice_buf_tbl = ice_find_buf_table(ice_seg);
+
+       ice_debug(hw, ICE_DBG_PKG, "Seg buf count: %d\n",
+                 LE32_TO_CPU(ice_buf_tbl->buf_count));
+
+       return ice_dwnld_cfg_bufs(hw, ice_buf_tbl->buf_array,
+                                 LE32_TO_CPU(ice_buf_tbl->buf_count));
+}
+
+/**
+ * ice_init_pkg_info
+ * @hw: pointer to the hardware structure
+ * @pkg_hdr: pointer to the driver's package hdr
+ *
+ * Saves off the package details into the hw structure.
+ */
+enum ice_status
+ice_init_pkg_info(struct ice_hw *hw, struct ice_pkg_hdr *pkg_hdr)
+{
+       struct ice_aqc_get_pkg_info_resp *pkg_info;
+       struct ice_global_metadata_seg *meta_seg;
+       struct ice_generic_seg_hdr *seg_hdr;
+       enum ice_status status;
+       u16 size;
+       u32 i;
+
+       ice_debug(hw, ICE_DBG_TRACE, "%s\n", __func__);
+       if (!pkg_hdr)
+               return ICE_ERR_PARAM;
+
+       meta_seg = (struct ice_global_metadata_seg *)
+                  ice_find_seg_in_pkg(hw, SEGMENT_TYPE_METADATA, pkg_hdr);
+       if (meta_seg) {
+               hw->pkg_ver = meta_seg->pkg_ver;
+               ice_memcpy(hw->pkg_name, meta_seg->pkg_name,
+                          sizeof(hw->pkg_name), ICE_NONDMA_TO_NONDMA);
+
+               ice_debug(hw, ICE_DBG_PKG, "Pkg: %d.%d.%d.%d, %s\n",
+                         meta_seg->pkg_ver.major, meta_seg->pkg_ver.minor,
+                         meta_seg->pkg_ver.update, meta_seg->pkg_ver.draft,
+                         meta_seg->pkg_name);
+       } else {
+               ice_debug(hw, ICE_DBG_INIT,
+                         "Did not find metadata segment in driver package\n");
+               return ICE_ERR_CFG;
+       }
+
+       seg_hdr = ice_find_seg_in_pkg(hw, SEGMENT_TYPE_ICE, pkg_hdr);
+       if (seg_hdr) {
+               hw->ice_pkg_ver = seg_hdr->seg_ver;
+               ice_memcpy(hw->ice_pkg_name, seg_hdr->seg_name,
+                          sizeof(hw->ice_pkg_name), ICE_NONDMA_TO_NONDMA);
+
+               ice_debug(hw, ICE_DBG_PKG, "Ice Pkg: %d.%d.%d.%d, %s\n",
+                         seg_hdr->seg_ver.major, seg_hdr->seg_ver.minor,
+                         seg_hdr->seg_ver.update, seg_hdr->seg_ver.draft,
+                         seg_hdr->seg_name);
+       } else {
+               ice_debug(hw, ICE_DBG_INIT,
+                         "Did not find ice segment in driver package\n");
+               return ICE_ERR_CFG;
+       }
+
+#define ICE_PKG_CNT    4
+       size = sizeof(*pkg_info) + (sizeof(pkg_info->pkg_info[0]) *
+                                   (ICE_PKG_CNT - 1));
+       pkg_info = (struct ice_aqc_get_pkg_info_resp *)ice_malloc(hw, size);
+       if (!pkg_info)
+               return ICE_ERR_NO_MEMORY;
+
+       status = ice_aq_get_pkg_info_list(hw, pkg_info, size, NULL);
+       if (status)
+               goto init_pkg_free_alloc;
+
+       for (i = 0; i < LE32_TO_CPU(pkg_info->count); i++) {
+#define ICE_PKG_FLAG_COUNT     4
+               char flags[ICE_PKG_FLAG_COUNT + 1] = { 0 };
+               u8 place = 0;
+
+               if (pkg_info->pkg_info[i].is_active) {
+                       flags[place++] = 'A';
+                       hw->active_pkg_ver = pkg_info->pkg_info[i].ver;
+                       ice_memcpy(hw->active_pkg_name,
+                                  pkg_info->pkg_info[i].name,
+                                  sizeof(hw->active_pkg_name),
+                                  ICE_NONDMA_TO_NONDMA);
+               }
+               if (pkg_info->pkg_info[i].is_active_at_boot)
+                       flags[place++] = 'B';
+               if (pkg_info->pkg_info[i].is_modified)
+                       flags[place++] = 'M';
+               if (pkg_info->pkg_info[i].is_in_nvm)
+                       flags[place++] = 'N';
+
+               ice_debug(hw, ICE_DBG_PKG, "Pkg[%d]: %d.%d.%d.%d,%s,%s\n",
+                         i, pkg_info->pkg_info[i].ver.major,
+                         pkg_info->pkg_info[i].ver.minor,
+                         pkg_info->pkg_info[i].ver.update,
+                         pkg_info->pkg_info[i].ver.draft,
+                         pkg_info->pkg_info[i].name, flags);
+       }
+
+init_pkg_free_alloc:
+       ice_free(hw, pkg_info);
+
+       return status;
+}
+
+/**
+ * ice_find_label_value
+ * @ice_seg: pointer to the ice segment (non-NULL)
+ * @name: name of the label to search for
+ * @type: the section type that will contain the label
+ * @value: pointer to a value that will return the label's value if found
+ *
+ * Finds a label's value given the label name and the section type to search.
+ * The ice_seg parameter must not be NULL since the first call to
+ * ice_enum_labels requires a pointer to an actual ice_seg structure.
+ */
+enum ice_status
+ice_find_label_value(struct ice_seg *ice_seg, char const *name, u32 type,
+                    u16 *value)
+{
+       struct ice_pkg_enum state;
+       char *label_name;
+       u16 val;
+
+       if (!ice_seg)
+               return ICE_ERR_PARAM;
+
+       do {
+               label_name = ice_enum_labels(ice_seg, type, &state, &val);
+               if (label_name && !strcmp(label_name, name)) {
+                       *value = val;
+                       return ICE_SUCCESS;
+               }
+
+               ice_seg = NULL;
+       } while (label_name);
+
+       return ICE_ERR_CFG;
+}
+
+/**
+ * ice_verify_pkg - verify package
+ * @pkg: pointer to the package buffer
+ * @len: size of the package buffer
+ *
+ * Verifies various attributes of the package file, including length, format
+ * version, and the requirement of at least one segment.
+ */
+static enum ice_status ice_verify_pkg(struct ice_pkg_hdr *pkg, u32 len)
+{
+       u32 seg_count;
+       u32 i;
+
+       if (len < sizeof(*pkg))
+               return ICE_ERR_BUF_TOO_SHORT;
+
+       if (pkg->format_ver.major != ICE_PKG_FMT_VER_MAJ ||
+           pkg->format_ver.minor != ICE_PKG_FMT_VER_MNR ||
+           pkg->format_ver.update != ICE_PKG_FMT_VER_UPD ||
+           pkg->format_ver.draft != ICE_PKG_FMT_VER_DFT)
+               return ICE_ERR_CFG;
+
+       /* pkg must have at least one segment */
+       seg_count = LE32_TO_CPU(pkg->seg_count);
+       if (seg_count < 1)
+               return ICE_ERR_CFG;
+
+       /* make sure segment array fits in package length */
+       if (len < sizeof(*pkg) + ((seg_count - 1) * sizeof(pkg->seg_offset)))
+               return ICE_ERR_BUF_TOO_SHORT;
+
+       /* all segments must fit within length */
+       for (i = 0; i < seg_count; i++) {
+               u32 off = LE32_TO_CPU(pkg->seg_offset[i]);
+               struct ice_generic_seg_hdr *seg;
+
+               /* segment header must fit */
+               if (len < off + sizeof(*seg))
+                       return ICE_ERR_BUF_TOO_SHORT;
+
+               seg = (struct ice_generic_seg_hdr *)((u8 *)pkg + off);
+
+               /* segment body must fit */
+               if (len < off + LE32_TO_CPU(seg->seg_size))
+                       return ICE_ERR_BUF_TOO_SHORT;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_free_seg - free package segment pointer
+ * @hw: pointer to the hardware structure
+ *
+ * Frees the package segment pointer in the proper manner, depending on if the
+ * segment was allocated or just the passed in pointer was stored.
+ */
+void ice_free_seg(struct ice_hw *hw)
+{
+       if (hw->pkg_copy) {
+               ice_free(hw, hw->pkg_copy);
+               hw->pkg_copy = NULL;
+       }
+       hw->seg = NULL;
+}
+
+/**
+ * ice_init_pkg_regs - initialize additional package registers
+ * @hw: pointer to the hardware structure
+ */
+static void ice_init_pkg_regs(struct ice_hw *hw)
+{
+#define ICE_SW_BLK_INP_MASK_L 0xFFFFFFFF
+#define ICE_SW_BLK_INP_MASK_H 0x0000FFFF
+#define ICE_SW_BLK_IDX 0
+
+       /* setup Switch block input mask, which is 48-bits in two parts */
+       wr32(hw, GL_PREEXT_L2_PMASK0(ICE_SW_BLK_IDX), ICE_SW_BLK_INP_MASK_L);
+       wr32(hw, GL_PREEXT_L2_PMASK1(ICE_SW_BLK_IDX), ICE_SW_BLK_INP_MASK_H);
+}
+
+/**
+ * ice_init_pkg - initialize/download package
+ * @hw: pointer to the hardware structure
+ * @buf: pointer to the package buffer
+ * @len: size of the package buffer
+ *
+ * This function initializes a package. The package contains HW tables
+ * required to do packet processing. First, the function extracts package
+ * information such as version. Then it finds the ice configuration segment
+ * within the package; this function then saves a copy of the segment pointer
+ * within the supplied package buffer. Next, the function will cache any hints
+ * from the package, followed by downloading the package itself. Note, that if
+ * a previous PF driver has already downloaded the package successfully, then
+ * the current driver will not have to download the package again.
+ *
+ * The local package contents will be used to query default behavior and to
+ * update specific sections of the HW's version of the package (e.g. to update
+ * the parse graph to understand new protocols).
+ *
+ * This function stores a pointer to the package buffer memory, and it is
+ * expected that the supplied buffer will not be freed immediately. If the
+ * package buffer needs to be freed, such as when read from a file, use
+ * ice_copy_and_init_pkg() instead of directly calling ice_init_pkg() in this
+ * case.
+ */
+static enum ice_status ice_init_pkg(struct ice_hw *hw, u8 *buf, u32 len)
+{
+       struct ice_pkg_hdr *pkg;
+       enum ice_status status;
+       struct ice_seg *seg;
+
+       if (!buf || !len)
+               return ICE_ERR_PARAM;
+
+       pkg = (struct ice_pkg_hdr *)buf;
+       status = ice_verify_pkg(pkg, len);
+       if (status) {
+               ice_debug(hw, ICE_DBG_INIT, "failed to verify pkg (err: %d)\n",
+                         status);
+               return status;
+       }
+
+       /* initialize package info */
+       status = ice_init_pkg_info(hw, pkg);
+       if (status)
+               return status;
+
+       /* find segment in given package */
+       seg = (struct ice_seg *)ice_find_seg_in_pkg(hw, SEGMENT_TYPE_ICE, pkg);
+       if (!seg) {
+               ice_debug(hw, ICE_DBG_INIT, "no ice segment in package.\n");
+               return ICE_ERR_CFG;
+       }
+
+       /* initialize package hints and then download package */
+       ice_init_pkg_hints(hw, seg);
+       status = ice_download_pkg(hw, seg);
+       if (status == ICE_ERR_AQ_NO_WORK) {
+               ice_debug(hw, ICE_DBG_INIT,
+                         "package previously loaded - no work.\n");
+               status = ICE_SUCCESS;
+       }
+
+       /* Free a previous segment, if necessary */
+       ice_free_seg(hw);
+       if (!status) {
+               hw->seg = seg;
+               /* on successful package download, update other required
+                * registers to support the package
+                */
+               ice_init_pkg_regs(hw);
+       } else {
+               ice_debug(hw, ICE_DBG_INIT, "package load failed, %d\n",
+                         status);
+       }
+
+       return status;
+}
+
+/**
+ * ice_copy_and_init_pkg - initialize/download a copy of the package
+ * @hw: pointer to the hardware structure
+ * @buf: pointer to the package buffer
+ * @len: size of the package buffer
+ *
+ * This function copies the package buffer, and then calls ice_init_pkg() to
+ * initialize the copied package contents.
+ *
+ * The copying is necessary if the package buffer supplied is constant, or if
+ * the memory may disappear shortly after calling this function.
+ *
+ * If the package buffer resides in the data segment and can be modified, the
+ * caller is free to use ice_init_pkg() instead of ice_copy_and_init_pkg().
+ *
+ * However, if the package buffer needs to be copied first, such as when being
+ * read from a file, the caller should use ice_copy_and_init_pkg().
+ *
+ * This function will first copy the package buffer, before calling
+ * ice_init_pkg(). The caller is free to immediately destroy the original
+ * package buffer, as the new copy will be managed by this function and
+ * related routines.
+ */
+enum ice_status ice_copy_and_init_pkg(struct ice_hw *hw, const u8 *buf, u32 len)
+{
+       enum ice_status status;
+       u8 *buf_copy;
+
+       if (!buf || !len)
+               return ICE_ERR_PARAM;
+
+       buf_copy = (u8 *)ice_memdup(hw, buf, len, ICE_NONDMA_TO_NONDMA);
+
+       status = ice_init_pkg(hw, buf_copy, len);
+       if (status)
+               /* Free the copy, since we failed to initialize the package */
+               ice_free(hw, buf_copy);
+       else
+               /* Track the copied pkg so we can free it later */
+               hw->pkg_copy = buf_copy;
+
+       return status;
+}
+
+/**
+ * ice_pkg_buf_alloc
+ * @hw: pointer to the HW structure
+ *
+ * Allocates a package buffer and returns a pointer to the buffer header.
+ * Note: all package contents must be in Little Endian form.
+ */
+struct ice_buf_build *ice_pkg_buf_alloc(struct ice_hw *hw)
+{
+       struct ice_buf_build *bld;
+       struct ice_buf_hdr *buf;
+
+       bld = (struct ice_buf_build *)ice_malloc(hw, sizeof(*bld));
+       if (!bld)
+               return NULL;
+
+       buf = (struct ice_buf_hdr *)bld;
+       buf->data_end = CPU_TO_LE16(sizeof(*buf) -
+                                   sizeof(buf->section_entry[0]));
+       return bld;
+}
+
+/**
+ * ice_sw_fv_handler
+ * @sect_type: section type
+ * @section: pointer to section
+ * @index: index of the field vector entry to be returned
+ * @offset: ptr to variable that receives the offset in the field vector table
+ *
+ * This is a callback function that can be passed to ice_pkg_enum_entry.
+ * This function treats the given section as of type ice_sw_fv_section and
+ * enumerates offset field. "offset" is an index into the field vector
+ * vector table.
+ */
+static void *
+ice_sw_fv_handler(u32 sect_type, void *section, u32 index, u32 *offset)
+{
+       struct ice_sw_fv_section *fv_section =
+               (struct ice_sw_fv_section *)section;
+
+       if (!section || sect_type != ICE_SID_FLD_VEC_SW)
+               return NULL;
+       if (index >= LE16_TO_CPU(fv_section->count))
+               return NULL;
+       if (offset)
+               /* "index" passed in to this function is relative to a given
+                * 4k block. To get to the true index into the field vector
+                * table need to add the relative index to the base_offset
+                * field of this section
+                */
+               *offset = LE16_TO_CPU(fv_section->base_offset) + index;
+       return fv_section->fv + index;
+}
+
+/**
+ * ice_get_sw_fv_list
+ * @hw: pointer to the HW structure
+ * @prot_ids: field vector to search for with a given protocol id
+ * @ids_cnt: lookup/protocol count
+ * @fv_list: Head of a list
+ *
+ * Finds all the field vector entries from switch block that contain
+ * a given protocol id and returns a list of structures of type
+ * "ice_sw_fv_list_entry". Every structure in the list has a field vector
+ * definition and profile id information
+ * NOTE: The caller of the function is responsible for freeing the memory
+ * allocated for every list entry.
+ */
+enum ice_status
+ice_get_sw_fv_list(struct ice_hw *hw, u16 *prot_ids, u8 ids_cnt,
+                  struct LIST_HEAD_TYPE *fv_list)
+{
+       struct ice_sw_fv_list_entry *fvl;
+       struct ice_sw_fv_list_entry *tmp;
+       struct ice_pkg_enum state;
+       struct ice_seg *ice_seg;
+       struct ice_fv *fv;
+       u32 offset;
+
+       if (!ids_cnt || !hw->seg)
+               return ICE_ERR_PARAM;
+
+       ice_seg = hw->seg;
+       do {
+               u8 i;
+
+               fv = (struct ice_fv *)
+                       ice_pkg_enum_entry(ice_seg, &state, ICE_SID_FLD_VEC_SW,
+                                          &offset, ice_sw_fv_handler);
+
+               for (i = 0; i < ids_cnt && fv; i++) {
+                       int j;
+
+                       /* This code assumes that if a switch field vector line
+                        * has a matching protocol, then this line will contain
+                        * the entries necessary to represent every field in
+                        * that protocol header.
+                        */
+                       for (j = 0; j < hw->blk[ICE_BLK_SW].es.fvw; j++)
+                               if (fv->ew[j].prot_id == prot_ids[i])
+                                       break;
+                       if (j >= hw->blk[ICE_BLK_SW].es.fvw)
+                               break;
+                       if (i + 1 == ids_cnt) {
+                               fvl = (struct ice_sw_fv_list_entry *)
+                                       ice_malloc(hw, sizeof(*fvl));
+                               if (!fvl)
+                                       goto err;
+                               fvl->fv_ptr = fv;
+                               fvl->profile_id = offset;
+                               LIST_ADD(&fvl->list_entry, fv_list);
+                               break;
+                       }
+               }
+               ice_seg = NULL;
+       } while (fv);
+       if (LIST_EMPTY(fv_list))
+               return ICE_ERR_CFG;
+       return ICE_SUCCESS;
+
+err:
+       LIST_FOR_EACH_ENTRY_SAFE(fvl, tmp, fv_list, ice_sw_fv_list_entry,
+                                list_entry) {
+               LIST_DEL(&fvl->list_entry);
+               ice_free(hw, fvl);
+       }
+
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_pkg_buf_alloc_single_section
+ * @hw: pointer to the HW structure
+ * @type: the section type value
+ * @size: the size of the section to reserve (in bytes)
+ * @section: returns pointer to the section
+ *
+ * Allocates a package buffer with a single section.
+ * Note: all package contents must be in Little Endian form.
+ */
+static struct ice_buf_build *
+ice_pkg_buf_alloc_single_section(struct ice_hw *hw, u32 type, u16 size,
+                                void **section)
+{
+       struct ice_buf_build *buf;
+
+       if (!section)
+               return NULL;
+
+       buf = ice_pkg_buf_alloc(hw);
+       if (!buf)
+               return NULL;
+
+       if (ice_pkg_buf_reserve_section(buf, 1))
+               goto ice_pkg_buf_alloc_single_section_err;
+
+       *section = ice_pkg_buf_alloc_section(buf, type, size);
+       if (!*section)
+               goto ice_pkg_buf_alloc_single_section_err;
+
+       return buf;
+
+ice_pkg_buf_alloc_single_section_err:
+       ice_pkg_buf_free(hw, buf);
+       return NULL;
+}
+
+/**
+ * ice_pkg_buf_reserve_section
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ * @count: the number of sections to reserve
+ *
+ * Reserves one or more section table entries in a package buffer. This routine
+ * can be called multiple times as long as they are made before calling
+ * ice_pkg_buf_alloc_section(). Once ice_pkg_buf_alloc_section()
+ * is called once, the number of sections that can be allocated will not be able
+ * to be increased; not using all reserved sections is fine, but this will
+ * result in some wasted space in the buffer.
+ * Note: all package contents must be in Little Endian form.
+ */
+enum ice_status
+ice_pkg_buf_reserve_section(struct ice_buf_build *bld, u16 count)
+{
+       struct ice_buf_hdr *buf;
+       u16 section_count;
+       u16 data_end;
+
+       if (!bld)
+               return ICE_ERR_PARAM;
+
+       buf = (struct ice_buf_hdr *)&bld->buf;
+
+       /* already an active section, can't increase table size */
+       section_count = LE16_TO_CPU(buf->section_count);
+       if (section_count > 0)
+               return ICE_ERR_CFG;
+
+       if (bld->reserved_section_table_entries + count > ICE_MAX_S_COUNT)
+               return ICE_ERR_CFG;
+       bld->reserved_section_table_entries += count;
+
+       data_end = LE16_TO_CPU(buf->data_end) +
+                  (count * sizeof(buf->section_entry[0]));
+       buf->data_end = CPU_TO_LE16(data_end);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_pkg_buf_unreserve_section
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ * @count: the number of sections to unreserve
+ *
+ * Unreserves one or more section table entries in a package buffer, releasing
+ * space that can be used for section data. This routine can be called
+ * multiple times as long as they are made before calling
+ * ice_pkg_buf_alloc_section(). Once ice_pkg_buf_alloc_section()
+ * is called once, the number of sections that can be allocated will not be able
+ * to be increased; not using all reserved sections is fine, but this will
+ * result in some wasted space in the buffer.
+ * Note: all package contents must be in Little Endian form.
+ */
+enum ice_status
+ice_pkg_buf_unreserve_section(struct ice_buf_build *bld, u16 count)
+{
+       struct ice_buf_hdr *buf;
+       u16 section_count;
+       u16 data_end;
+
+       if (!bld)
+               return ICE_ERR_PARAM;
+
+       buf = (struct ice_buf_hdr *)&bld->buf;
+
+       /* already an active section, can't decrease table size */
+       section_count = LE16_TO_CPU(buf->section_count);
+       if (section_count > 0)
+               return ICE_ERR_CFG;
+
+       if (count > bld->reserved_section_table_entries)
+               return ICE_ERR_CFG;
+       bld->reserved_section_table_entries -= count;
+
+       data_end = LE16_TO_CPU(buf->data_end) -
+                  (count * sizeof(buf->section_entry[0]));
+       buf->data_end = CPU_TO_LE16(data_end);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_pkg_buf_alloc_section
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ * @type: the section type value
+ * @size: the size of the section to reserve (in bytes)
+ *
+ * Reserves memory in the buffer for a section's content and updates the
+ * buffers' status accordingly. This routine returns a pointer to the first
+ * byte of the section start within the buffer, which is used to fill in the
+ * section contents.
+ * Note: all package contents must be in Little Endian form.
+ */
+void *
+ice_pkg_buf_alloc_section(struct ice_buf_build *bld, u32 type, u16 size)
+{
+       struct ice_buf_hdr *buf;
+       u16 sect_count;
+       u16 data_end;
+
+       if (!bld || !type || !size)
+               return NULL;
+
+       buf = (struct ice_buf_hdr *)&bld->buf;
+
+       /* check for enough space left in buffer */
+       data_end = LE16_TO_CPU(buf->data_end);
+
+       /* section start must align on 4 byte boundary */
+       data_end = ICE_ALIGN(data_end, 4);
+
+       if ((data_end + size) > ICE_MAX_S_DATA_END)
+               return NULL;
+
+       /* check for more available section table entries */
+       sect_count = LE16_TO_CPU(buf->section_count);
+       if (sect_count < bld->reserved_section_table_entries) {
+               void *section_ptr = ((u8 *)buf) + data_end;
+
+               buf->section_entry[sect_count].offset = CPU_TO_LE16(data_end);
+               buf->section_entry[sect_count].size = CPU_TO_LE16(size);
+               buf->section_entry[sect_count].type = CPU_TO_LE32(type);
+
+               data_end += size;
+               buf->data_end = CPU_TO_LE16(data_end);
+
+               buf->section_count = CPU_TO_LE16(sect_count + 1);
+               return section_ptr;
+       }
+
+       /* no free section table entries */
+       return NULL;
+}
+
+/**
+ * ice_pkg_buf_get_free_space
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ *
+ * Returns the number of free bytes remaining in the buffer.
+ * Note: all package contents must be in Little Endian form.
+ */
+u16 ice_pkg_buf_get_free_space(struct ice_buf_build *bld)
+{
+       struct ice_buf_hdr *buf;
+
+       if (!bld)
+               return 0;
+
+       buf = (struct ice_buf_hdr *)&bld->buf;
+       return ICE_MAX_S_DATA_END - LE16_TO_CPU(buf->data_end);
+}
+
+/**
+ * ice_pkg_buf_get_active_sections
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ *
+ * Returns the number of active sections. Before using the package buffer
+ * in an update package command, the caller should make sure that there is at
+ * least one active section - otherwise, the buffer is not legal and should
+ * not be used.
+ * Note: all package contents must be in Little Endian form.
+ */
+u16 ice_pkg_buf_get_active_sections(struct ice_buf_build *bld)
+{
+       struct ice_buf_hdr *buf;
+
+       if (!bld)
+               return 0;
+
+       buf = (struct ice_buf_hdr *)&bld->buf;
+       return LE16_TO_CPU(buf->section_count);
+}
+
+/**
+ * ice_pkg_buf_header
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ *
+ * Return a pointer to the buffer's header
+ */
+struct ice_buf *ice_pkg_buf(struct ice_buf_build *bld)
+{
+       if (!bld)
+               return NULL;
+
+       return &bld->buf;
+}
+
+/**
+ * ice_pkg_buf_free
+ * @hw: pointer to the HW structure
+ * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc())
+ *
+ * Frees a package buffer
+ */
+void ice_pkg_buf_free(struct ice_hw *hw, struct ice_buf_build *bld)
+{
+       ice_free(hw, bld);
+}
+
+/* PTG Management */
+
+/**
+ * ice_ptg_update_xlt1 - Updates packet type groups in hw via xlt1 table
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ *
+ * This function will update the xlt1 hardware table to reflect the new
+ * packet type group configuration.
+ */
+enum ice_status ice_ptg_update_xlt1(struct ice_hw *hw, enum ice_block blk)
+{
+       struct ice_xlt1_section *sect;
+       struct ice_buf_build *bld;
+       enum ice_status status;
+       u16 index;
+
+       bld = ice_pkg_buf_alloc_single_section(hw, ice_sect_id(blk, ICE_XLT1),
+                                              ICE_XLT1_SIZE(ICE_XLT1_CNT),
+                                              (void **)&sect);
+       if (!bld)
+               return ICE_ERR_NO_MEMORY;
+
+       sect->count = CPU_TO_LE16(ICE_XLT1_CNT);
+       sect->offset = CPU_TO_LE16(0);
+       for (index = 0; index < ICE_XLT1_CNT; index++)
+               sect->value[index] = hw->blk[blk].xlt1.ptypes[index].ptg;
+
+       status = ice_update_pkg(hw, ice_pkg_buf(bld), 1);
+
+       ice_pkg_buf_free(hw, bld);
+
+       return status;
+}
+
+/**
+ * ice_ptg_find_ptype - Search for packet type group using packet type (ptype)
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptype: the ptype to search for
+ * @ptg: pointer to variable that receives the PTG
+ *
+ * This function will search the PTGs for a particular ptype, returning the
+ * PTG ID that contains it through the ptg parameter, with the value of
+ * ICE_DEFAULT_PTG (0) meaning it is part the default PTG.
+ */
+enum ice_status
+ice_ptg_find_ptype(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 *ptg)
+{
+       if (ptype >= ICE_XLT1_CNT || !ptg)
+               return ICE_ERR_PARAM;
+
+       *ptg = hw->blk[blk].xlt1.ptypes[ptype].ptg;
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_ptg_alloc_val - Allocates a new packet type group ID by value
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptg: the ptg to allocate
+ *
+ * This function allocates a given packet type group ID specified by the ptg
+ * parameter.
+ */
+static
+void ice_ptg_alloc_val(struct ice_hw *hw, enum ice_block blk, u8 ptg)
+{
+       hw->blk[blk].xlt1.ptg_tbl[ptg].in_use = true;
+}
+
+/**
+ * ice_ptg_alloc - Find a free entry and allocates a new packet type group ID
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ *
+ * This function allocates and returns a new packet type group ID. Note
+ * that 0 is the default packet type group, so successfully created PTGs will
+ * have a non-zero ID value; which means a 0 return value indicates an error.
+ */
+u8 ice_ptg_alloc(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 i;
+
+       /* Skip the default PTG of 0 */
+       for (i = 1; i < ICE_MAX_PTGS; i++)
+               if (!hw->blk[blk].xlt1.ptg_tbl[i].in_use) {
+                       /* found a free PTG ID */
+                       ice_ptg_alloc_val(hw, blk, i);
+                       return (u8)i;
+               }
+
+       return 0;
+}
+
+/**
+ * ice_ptg_free - Frees a packet type group
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptg: the ptg ID to free
+ *
+ * This function frees a packet type group, and returns all the current ptypes
+ * within it to the default PTG.
+ */
+void ice_ptg_free(struct ice_hw *hw, enum ice_block blk, u8 ptg)
+{
+       struct ice_ptg_ptype *p, *temp;
+
+       hw->blk[blk].xlt1.ptg_tbl[ptg].in_use = false;
+       p = hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype;
+       while (p) {
+               p->ptg = ICE_DEFAULT_PTG;
+               temp = p->next_ptype;
+               p->next_ptype = NULL;
+               p = temp;
+       }
+
+       hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype = NULL;
+}
+
+/**
+ * ice_ptg_remove_ptype - Removes ptype from a particular packet type group
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptype: the ptype to remove
+ * @ptg: the ptg to remove the ptype from
+ *
+ * This function will remove the ptype from the specific ptg, and move it to
+ * the default PTG (ICE_DEFAULT_PTG).
+ */
+static enum ice_status
+ice_ptg_remove_ptype(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 ptg)
+{
+       struct ice_ptg_ptype **ch;
+       struct ice_ptg_ptype *p;
+
+       if (ptype > ICE_XLT1_CNT - 1)
+               return ICE_ERR_PARAM;
+
+       if (!hw->blk[blk].xlt1.ptg_tbl[ptg].in_use)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       /* Should not happen if .in_use is set, bad config */
+       if (!hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype)
+               return ICE_ERR_CFG;
+
+       /* find the ptype within this PTG, and bypass the link over it */
+       p = hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype;
+       ch = &hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype;
+       while (p) {
+               if (ptype == (p - hw->blk[blk].xlt1.ptypes)) {
+                       *ch = p->next_ptype;
+                       break;
+               }
+
+               ch = &p->next_ptype;
+               p = p->next_ptype;
+       }
+
+       hw->blk[blk].xlt1.ptypes[ptype].ptg = ICE_DEFAULT_PTG;
+       hw->blk[blk].xlt1.ptypes[ptype].next_ptype = NULL;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_ptg_add_mv_ptype - Adds/moves ptype to a particular packet type group
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptype: the ptype to add or move
+ * @ptg: the ptg to add or move the ptype to
+ *
+ * This function will either add or move a ptype to a particular PTG depending
+ * on if the ptype is already part of another group. Note that using a
+ * a destination PTG ID of ICE_DEFAULT_PTG (0) will move the ptype to the
+ * default PTG.
+ */
+enum ice_status
+ice_ptg_add_mv_ptype(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 ptg)
+{
+       enum ice_status status;
+       u8 original_ptg;
+
+       if (ptype > ICE_XLT1_CNT - 1)
+               return ICE_ERR_PARAM;
+
+       if (!hw->blk[blk].xlt1.ptg_tbl[ptg].in_use && ptg != ICE_DEFAULT_PTG)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       status = ice_ptg_find_ptype(hw, blk, ptype, &original_ptg);
+       if (status)
+               return status;
+
+       /* Is ptype already in the correct PTG? */
+       if (original_ptg == ptg)
+               return ICE_SUCCESS;
+
+       /* Remove from original PTG and move back to the default PTG */
+       if (original_ptg != ICE_DEFAULT_PTG)
+               ice_ptg_remove_ptype(hw, blk, ptype, original_ptg);
+
+       /* Moving to default PTG? Then we're done with this request */
+       if (ptg == ICE_DEFAULT_PTG)
+               return ICE_SUCCESS;
+
+       /* Add ptype to PTG at beginning of list */
+       hw->blk[blk].xlt1.ptypes[ptype].next_ptype =
+               hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype;
+       hw->blk[blk].xlt1.ptg_tbl[ptg].first_ptype =
+               &hw->blk[blk].xlt1.ptypes[ptype];
+
+       hw->blk[blk].xlt1.ptypes[ptype].ptg = ptg;
+       hw->blk[blk].xlt1.t[ptype] = ptg;
+
+       return ICE_SUCCESS;
+}
+
+/* Block / table size info */
+struct ice_blk_size_details {
+       u16 xlt1;                       /* # xlt1 entries */
+       u16 xlt2;                       /* # xlt2 entries */
+       u16 prof_tcam;                  /* # profile id tcam entries */
+       u16 prof_id;                    /* # profile ids */
+       u8 prof_cdid_bits;              /* # cdid one-hot bits used in key */
+       u16 prof_redir;                 /* # profile redirection entries */
+       u16 es;                         /* # extraction sequence entries */
+       u16 fvw;                        /* # field vector words */
+       u8 overwrite;                   /* overwrite existing entries allowed */
+       u8 reverse;                     /* reverse FV order */
+};
+
+static const struct ice_blk_size_details blk_sizes[ICE_BLK_COUNT] = {
+       /**
+        * Table Definitions
+        * XLT1 - Number of entries in XLT1 table
+        * XLT2 - Number of entries in XLT2 table
+        * TCAM - Number of entries Profile ID TCAM table
+        * CDID - Control Domain ID of the hardware block
+        * PRED - Number of entries in the Profile Redirection Table
+        * FV   - Number of entries in the Field Vector
+        * FVW  - Width (in WORDs) of the Field Vector
+        * OVR  - Overwrite existing table entries
+        * REV  - Reverse FV
+        */
+       /*          XLT1        , XLT2        ,TCAM, PID,CDID,PRED,   FV, FVW */
+       /*          Overwrite   , Reverse FV */
+       /* SW  */ { ICE_XLT1_CNT, ICE_XLT2_CNT, 512, 256,   0,  256, 256,  48,
+                   false, false },
+       /* ACL */ { ICE_XLT1_CNT, ICE_XLT2_CNT, 512, 128,   0,  128, 128,  32,
+                   false, false },
+       /* FD  */ { ICE_XLT1_CNT, ICE_XLT2_CNT, 512, 128,   0,  128, 128,  24,
+                   false, true  },
+       /* RSS */ { ICE_XLT1_CNT, ICE_XLT2_CNT, 512, 128,   0,  128, 128,  24,
+                   true,  true  },
+       /* PE  */ { ICE_XLT1_CNT, ICE_XLT2_CNT,  64,  32,   0,   32,  32,  24,
+                   false, false },
+};
+
+enum ice_sid_all {
+       ICE_SID_XLT1_OFF = 0,
+       ICE_SID_XLT2_OFF,
+       ICE_SID_PR_OFF,
+       ICE_SID_PR_REDIR_OFF,
+       ICE_SID_ES_OFF,
+       ICE_SID_OFF_COUNT,
+};
+
+/* Characteristic handling */
+
+/**
+ * ice_match_prop_lst - determine if properties of two lists match
+ * @list1: first properties list
+ * @list2: second properties list
+ *
+ * Count, cookies and the order must match in order to be considered equivalent.
+ */
+static bool
+ice_match_prop_lst(struct LIST_HEAD_TYPE *list1, struct LIST_HEAD_TYPE *list2)
+{
+       struct ice_vsig_prof *tmp1;
+       struct ice_vsig_prof *tmp2;
+       u16 chk_count = 0;
+       u16 count = 0;
+
+       /* compare counts */
+       LIST_FOR_EACH_ENTRY(tmp1, list1, ice_vsig_prof, list) {
+               count++;
+       }
+       LIST_FOR_EACH_ENTRY(tmp2, list2, ice_vsig_prof, list) {
+               chk_count++;
+       }
+       if (!count || count != chk_count)
+               return false;
+
+       tmp1 = LIST_FIRST_ENTRY(list1, struct ice_vsig_prof, list);
+       tmp2 = LIST_FIRST_ENTRY(list2, struct ice_vsig_prof, list);
+
+       /* profile cookies must compare, and in the exact same order to take
+        * into account priority
+        */
+       while (--count) {
+               if (tmp2->profile_cookie != tmp1->profile_cookie)
+                       return false;
+
+               tmp1 = LIST_NEXT_ENTRY(tmp1, struct ice_vsig_prof, list);
+               tmp2 = LIST_NEXT_ENTRY(tmp2, struct ice_vsig_prof, list);
+       }
+
+       return true;
+}
+
+/* VSIG Management */
+
+/**
+ * ice_vsig_update_xlt2_sect - update one section of xlt2 table
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsi: hw vsi number to program
+ * @vsig: vsig for the vsi
+ *
+ * This function will update the xlt2 hardware table with the input vsi
+ * group configuration.
+ */
+static enum ice_status
+ice_vsig_update_xlt2_sect(struct ice_hw *hw, enum ice_block blk, u16 vsi,
+                         u16 vsig)
+{
+       struct ice_xlt2_section *sect;
+       struct ice_buf_build *bld;
+       enum ice_status status;
+
+       bld = ice_pkg_buf_alloc_single_section(hw, ice_sect_id(blk, ICE_XLT2),
+                                              sizeof(struct ice_xlt2_section),
+                                              (void **)&sect);
+       if (!bld)
+               return ICE_ERR_NO_MEMORY;
+
+       sect->count = CPU_TO_LE16(1);
+       sect->offset = CPU_TO_LE16(vsi);
+       sect->value[0] = CPU_TO_LE16(vsig);
+
+       status = ice_update_pkg(hw, ice_pkg_buf(bld), 1);
+
+       ice_pkg_buf_free(hw, bld);
+
+       return status;
+}
+
+/**
+ * ice_vsig_update_xlt2 - update xlt2 table with VSIG configuration
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ *
+ * This function will update the xlt2 hardware table with the input vsi
+ * group configuration of used vsis.
+ */
+enum ice_status ice_vsig_update_xlt2(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 vsi;
+
+       for (vsi = 0; vsi < ICE_MAX_VSI; vsi++) {
+               /* update only vsis that have been changed */
+               if (hw->blk[blk].xlt2.vsis[vsi].changed) {
+                       enum ice_status status;
+                       u16 vsig;
+
+                       vsig = hw->blk[blk].xlt2.vsis[vsi].vsig;
+                       status = ice_vsig_update_xlt2_sect(hw, blk, vsi, vsig);
+                       if (status)
+                               return status;
+
+                       hw->blk[blk].xlt2.vsis[vsi].changed = 0;
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_vsig_find_vsi - find a VSIG that contains a specified vsi
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsi: vsi of interest
+ * @vsig: pointer to receive the vsi group
+ *
+ * This function will lookup the vsi entry in the XLT2 list and return
+ * the vsi group its associated with.
+ */
+enum ice_status
+ice_vsig_find_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 *vsig)
+{
+       if (!vsig || vsi >= ICE_MAX_VSI)
+               return ICE_ERR_PARAM;
+
+       /* As long as there's a default or valid VSIG associated with the input
+        * vsi, the functions returns a success. Any handling of VSIG will be
+        * done by the following add, update or remove functions.
+        */
+       *vsig = hw->blk[blk].xlt2.vsis[vsi].vsig;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_vsig_alloc_val - allocate a new VSIG by value
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsig: the vsig to allocate
+ *
+ * This function will allocate a given VSIG specified by the vsig parameter.
+ */
+static u16 ice_vsig_alloc_val(struct ice_hw *hw, enum ice_block blk, u16 vsig)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+
+       if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use) {
+               INIT_LIST_HEAD(&hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst);
+               hw->blk[blk].xlt2.vsig_tbl[idx].in_use = true;
+       }
+
+       return ICE_VSIG_VALUE(idx, hw->pf_id);
+}
+
+/**
+ * ice_vsig_alloc - Finds a free entry and allocates a new VSIG
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ *
+ * This function will iterate through the VSIG list and mark the first
+ * unused entry for the new VSIG entry as used and return that value.
+ */
+static u16 ice_vsig_alloc(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 i;
+
+       for (i = 1; i < ICE_MAX_VSIGS; i++)
+               if (!hw->blk[blk].xlt2.vsig_tbl[i].in_use)
+                       return ice_vsig_alloc_val(hw, blk, i);
+
+       return ICE_DEFAULT_VSIG;
+}
+
+/**
+ * ice_find_dup_props_vsig - find vsi group with a specified set of properties
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @chs: characteristic list
+ * @vsig: returns the VSIG with the matching profiles, if found
+ *
+ * Each VSIG is associated with a characteristic set; i.e. all vsis under
+ * a group have the same characteristic set. To check if there exists a VSIG
+ * which has the same characteristics as the input characteristics; this
+ * function will iterate through the xlt2 list and return the VSIG that has a
+ * matching configuration. In order to make sure that priorities are accounted
+ * for, the list must match exactly, including the order in which the
+ * characteristics are listed.
+ */
+enum ice_status
+ice_find_dup_props_vsig(struct ice_hw *hw, enum ice_block blk,
+                       struct LIST_HEAD_TYPE *chs, u16 *vsig)
+{
+       struct ice_xlt2 *xlt2 = &hw->blk[blk].xlt2;
+       u16 i;
+
+       for (i = 0; i < xlt2->count; i++) {
+               if (xlt2->vsig_tbl[i].in_use &&
+                   ice_match_prop_lst(chs, &xlt2->vsig_tbl[i].prop_lst)) {
+                       *vsig = (i | ((hw->pf_id << ICE_PF_NUM_S) &
+                                     ICE_PF_NUM_M));
+                       *vsig = ICE_VSIG_VALUE(i, hw->pf_id);
+                       return ICE_SUCCESS;
+               }
+       }
+
+       return ICE_ERR_DOES_NOT_EXIST;
+}
+
+/**
+ * ice_vsig_free - free vsi group
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsig: VSIG to remove
+ *
+ * The function will remove all vsis associated with the input VSIG and move
+ * them to the DEFAULT_VSIG and mark the VSIG available.
+ */
+enum ice_status
+ice_vsig_free(struct ice_hw *hw, enum ice_block blk, u16 vsig)
+{
+       struct ice_vsig_prof *dtmp, *del;
+       struct ice_vsig_vsi *vsi_cur;
+       u16 idx;
+
+       idx = vsig & ICE_VSIG_IDX_M;
+       if (idx >= ICE_MAX_VSIGS)
+               return ICE_ERR_PARAM;
+
+       if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       hw->blk[blk].xlt2.vsig_tbl[idx].in_use = false;
+
+       vsi_cur = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi;
+       if (!vsi_cur)
+               return ICE_ERR_CFG;
+
+       /* remove all vsis associated with this VSIG XLT2 entry */
+       do {
+               struct ice_vsig_vsi *tmp = vsi_cur->next_vsi;
+
+               vsi_cur->vsig = ICE_DEFAULT_VSIG;
+               vsi_cur->changed = 1;
+               vsi_cur->next_vsi = NULL;
+               vsi_cur = tmp;
+       } while (vsi_cur);
+
+       /* NULL terminate head of vsi list */
+       hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi = NULL;
+
+       /* free characteristic list */
+       LIST_FOR_EACH_ENTRY_SAFE(del, dtmp,
+                                &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                                ice_vsig_prof, list) {
+               LIST_DEL(&del->list);
+               ice_free(hw, del);
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_vsig_add_mv_vsi - add or move a vsi to a vsi group
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsi: vsi to move
+ * @vsig: destination vsi group
+ *
+ * This function will move or add the input vsi to the target VSIG.
+ * The function will find the original VSIG the vsi belongs to and
+ * move the entry to the DEFAULT_VSIG, update the original VSIG and
+ * then move entry to the new VSIG.
+ */
+enum ice_status
+ice_vsig_add_mv_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig)
+{
+       struct ice_vsig_vsi *tmp;
+       enum ice_status status;
+       u16 orig_vsig, idx;
+
+       idx = vsig & ICE_VSIG_IDX_M;
+
+       if (vsi >= ICE_MAX_VSI || idx >= ICE_MAX_VSIGS)
+               return ICE_ERR_PARAM;
+
+       /* if VSIG not in use and VSIG is not default type this VSIG
+        * doesn't exist.
+        */
+       if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use &&
+           vsig != ICE_DEFAULT_VSIG)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       status = ice_vsig_find_vsi(hw, blk, vsi, &orig_vsig);
+       if (status)
+               return status;
+
+       /* no update required if vsigs match */
+       if (orig_vsig == vsig)
+               return ICE_SUCCESS;
+
+       if (orig_vsig != ICE_DEFAULT_VSIG) {
+               /* remove entry from orig_vsig and add to default VSIG */
+               status = ice_vsig_remove_vsi(hw, blk, vsi, orig_vsig);
+               if (status)
+                       return status;
+       }
+
+       if (idx == ICE_DEFAULT_VSIG)
+               return ICE_SUCCESS;
+
+       /* Create vsi entry and add VSIG and prop_mask values */
+       hw->blk[blk].xlt2.vsis[vsi].vsig = vsig;
+       hw->blk[blk].xlt2.vsis[vsi].changed = 1;
+
+       /* Add new entry to the head of the VSIG list */
+       tmp = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi;
+       hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi =
+               &hw->blk[blk].xlt2.vsis[vsi];
+       hw->blk[blk].xlt2.vsis[vsi].next_vsi = tmp;
+       hw->blk[blk].xlt2.t[vsi] = vsig;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_vsig_remove_vsi - remove vsi from VSIG
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsi: vsi to remove
+ * @vsig: vsi group to remove from
+ *
+ * The function will remove the input vsi from its vsi group and move it
+ * to the DEFAULT_VSIG.
+ */
+enum ice_status
+ice_vsig_remove_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig)
+{
+       struct ice_vsig_vsi **vsi_head, *vsi_cur, *vsi_tgt;
+       u16 idx;
+
+       idx = vsig & ICE_VSIG_IDX_M;
+
+       if (vsi >= ICE_MAX_VSI || idx >= ICE_MAX_VSIGS)
+               return ICE_ERR_PARAM;
+
+       if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       /* entry already in default VSIG, dont have to remove */
+       if (idx == ICE_DEFAULT_VSIG)
+               return ICE_SUCCESS;
+
+       vsi_head = &hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi;
+       if (!(*vsi_head))
+               return ICE_ERR_CFG;
+
+       vsi_tgt = &hw->blk[blk].xlt2.vsis[vsi];
+       vsi_cur = (*vsi_head);
+
+       /* iterate the vsi list, skip over the entry to be removed */
+       while (vsi_cur) {
+               if (vsi_tgt == vsi_cur) {
+                       (*vsi_head) = vsi_cur->next_vsi;
+                       break;
+               }
+               vsi_head = &vsi_cur->next_vsi;
+               vsi_cur = vsi_cur->next_vsi;
+       }
+
+       /* verify if vsi was removed from group list */
+       if (!vsi_cur)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       vsi_cur->vsig = ICE_DEFAULT_VSIG;
+       vsi_cur->changed = 1;
+       vsi_cur->next_vsi = NULL;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_find_prof_id - find profile id for a given field vector
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @fv: field vector to search for
+ * @prof_id: receives the profile id
+ */
+static enum ice_status
+ice_find_prof_id(struct ice_hw *hw, enum ice_block blk,
+                struct ice_fv_word *fv, u8 *prof_id)
+{
+       struct ice_es *es = &hw->blk[blk].es;
+       u16 off, i;
+
+       for (i = 0; i < es->count; i++) {
+               off = i * es->fvw;
+
+               if (memcmp(&es->t[off], fv, es->fvw * 2))
+                       continue;
+
+               *prof_id = i;
+               return ICE_SUCCESS;
+       }
+
+       return ICE_ERR_DOES_NOT_EXIST;
+}
+
+/**
+ * ice_prof_id_rsrc_type - get profile id resource type for a block type
+ * @blk: the block type
+ * @rsrc_type: pointer to variable to receive the resource type
+ */
+static bool ice_prof_id_rsrc_type(enum ice_block blk, u16 *rsrc_type)
+{
+       switch (blk) {
+       case ICE_BLK_SW:
+               *rsrc_type = ICE_AQC_RES_TYPE_SWITCH_PROF_BLDR_PROFID;
+               break;
+       case ICE_BLK_ACL:
+               *rsrc_type = ICE_AQC_RES_TYPE_ACL_PROF_BLDR_PROFID;
+               break;
+       case ICE_BLK_FD:
+               *rsrc_type = ICE_AQC_RES_TYPE_FD_PROF_BLDR_PROFID;
+               break;
+       case ICE_BLK_RSS:
+               *rsrc_type = ICE_AQC_RES_TYPE_HASH_PROF_BLDR_PROFID;
+               break;
+       case ICE_BLK_PE:
+               *rsrc_type = ICE_AQC_RES_TYPE_QHASH_PROF_BLDR_PROFID;
+               break;
+       default:
+               return false;
+       }
+       return true;
+}
+
+/**
+ * ice_tcam_ent_rsrc_type - get tcam entry resource type for a block type
+ * @blk: the block type
+ * @rsrc_type: pointer to variable to receive the resource type
+ */
+static bool ice_tcam_ent_rsrc_type(enum ice_block blk, u16 *rsrc_type)
+{
+       switch (blk) {
+       case ICE_BLK_SW:
+               *rsrc_type = ICE_AQC_RES_TYPE_SWITCH_PROF_BLDR_TCAM;
+               break;
+       case ICE_BLK_ACL:
+               *rsrc_type = ICE_AQC_RES_TYPE_ACL_PROF_BLDR_TCAM;
+               break;
+       case ICE_BLK_FD:
+               *rsrc_type = ICE_AQC_RES_TYPE_FD_PROF_BLDR_TCAM;
+               break;
+       case ICE_BLK_RSS:
+               *rsrc_type = ICE_AQC_RES_TYPE_HASH_PROF_BLDR_TCAM;
+               break;
+       case ICE_BLK_PE:
+               *rsrc_type = ICE_AQC_RES_TYPE_QHASH_PROF_BLDR_TCAM;
+               break;
+       default:
+               return false;
+       }
+       return true;
+}
+
+/**
+ * ice_workaround_get_res_blk - determine the block from a resource type
+ * @type: type of resource
+ * @blk: pointer to a enum that will receive the block type
+ * @tcam: pointer to variable that will be set to true for a TCAM resource type
+ */
+static enum
+ice_status ice_workaround_get_res_blk(u16 type, enum ice_block *blk, bool *tcam)
+{
+       /* just need to support TCAM entries and Profile IDs for now */
+       *tcam = false;
+
+       switch (type) {
+       case ICE_AQC_RES_TYPE_SWITCH_PROF_BLDR_TCAM:
+               *blk = ICE_BLK_SW;
+               *tcam = true;
+               break;
+       case ICE_AQC_RES_TYPE_ACL_PROF_BLDR_TCAM:
+               *blk = ICE_BLK_ACL;
+               *tcam = true;
+               break;
+       case ICE_AQC_RES_TYPE_FD_PROF_BLDR_TCAM:
+               *blk = ICE_BLK_FD;
+               *tcam = true;
+               break;
+       case ICE_AQC_RES_TYPE_HASH_PROF_BLDR_TCAM:
+               *blk = ICE_BLK_RSS;
+               *tcam = true;
+               break;
+       case ICE_AQC_RES_TYPE_QHASH_PROF_BLDR_TCAM:
+               *blk = ICE_BLK_PE;
+               *tcam = true;
+               break;
+       case ICE_AQC_RES_TYPE_SWITCH_PROF_BLDR_PROFID:
+               *blk = ICE_BLK_SW;
+               break;
+       case ICE_AQC_RES_TYPE_ACL_PROF_BLDR_PROFID:
+               *blk = ICE_BLK_ACL;
+               break;
+       case ICE_AQC_RES_TYPE_FD_PROF_BLDR_PROFID:
+               *blk = ICE_BLK_FD;
+               break;
+       case ICE_AQC_RES_TYPE_HASH_PROF_BLDR_PROFID:
+               *blk = ICE_BLK_RSS;
+               break;
+       case ICE_AQC_RES_TYPE_QHASH_PROF_BLDR_PROFID:
+               *blk = ICE_BLK_PE;
+               break;
+       default:
+               return ICE_ERR_PARAM;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_alloc_res_workaround
+ * @hw: pointer to the hw struct
+ * @type: type of resource
+ * @num: number of resources to allocate
+ * @res: pointer to array that will receive the resources
+ */
+static enum ice_status
+ice_alloc_res_workaround(struct ice_hw *hw, u16 type, u16 num, u16 *res)
+{
+       enum ice_block blk;
+       u16 count = 0;
+       bool tcam;
+       u16 first;
+       u16 last;
+       u16 max;
+       u16 i;
+
+/* Number of PFs we support with this workaround */
+#define ICE_WA_PF_COUNT        4
+#define ICE_WA_1ST_TCAM        4
+#define ICE_WA_1ST_FV  4
+
+       /* Only allow our supported PFs */
+       if (hw->pf_id >= ICE_WA_PF_COUNT)
+               return ICE_ERR_AQ_ERROR;
+
+       if (ice_workaround_get_res_blk(type, &blk, &tcam))
+               return ICE_ERR_AQ_ERROR;
+
+       if (tcam) {
+               /* range of entries based on PF */
+               max = hw->blk[blk].prof.count / ICE_WA_PF_COUNT;
+               first = max * hw->pf_id;
+               last = first + max;
+
+               /* Profile IDs - start at non-zero index for PROF ID TCAM table
+                * The first few entries are for bypass, default and errors
+                * (only relevant for PF 0)
+                */
+               first += hw->pf_id ? 0 : ICE_WA_1ST_TCAM;
+
+               for (i = first; i < last && count < num; i++) {
+                       if (!hw->blk[blk].prof.resource_used_hack[i]) {
+                               res[count++] = i;
+                               hw->blk[blk].prof.resource_used_hack[i] = true;
+                       }
+               }
+
+               /* handle failure case */
+               if (count < num) {
+                       for (i = 0; i < count; i++) {
+                               hw->blk[blk].prof.resource_used_hack[res[i]] =
+                                       false;
+                               res[i] = 0;
+                       }
+
+                       return ICE_ERR_AQ_ERROR;
+               }
+       } else {
+               /* range of entries based on PF */
+               max = hw->blk[blk].es.count / ICE_WA_PF_COUNT;
+               first = max * hw->pf_id;
+               last = first + max;
+
+               /* FV index - start at non-zero index for Field vector table
+                * The first few entries are for bypass, default and errors
+                * (only relevant for PF 0)
+                */
+               first += hw->pf_id ? 0 : ICE_WA_1ST_FV;
+
+               for (i = first; i < last && count < num; i++) {
+                       if (!hw->blk[blk].es.resource_used_hack[i]) {
+                               res[count++] = i;
+                               hw->blk[blk].es.resource_used_hack[i] = true;
+                       }
+               }
+
+               /* handle failure case */
+               if (count < num) {
+                       for (i = 0; i < count; i++) {
+                               hw->blk[blk].es.resource_used_hack[res[i]] =
+                                       false;
+                               res[i] = 0;
+                       }
+
+                       return ICE_ERR_AQ_ERROR;
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_free_res_workaround
+ * @hw: pointer to the hw struct
+ * @type: type of resource to free
+ * @num: number of resources
+ * @res: array of resource ids to free
+ */
+static enum ice_status
+ice_free_res_workaround(struct ice_hw *hw, u16 type, u16 num, u16 *res)
+{
+       enum ice_block blk;
+       bool tcam = false;
+       u16 i;
+
+       if (ice_workaround_get_res_blk(type, &blk, &tcam))
+               return ICE_ERR_AQ_ERROR;
+
+       if (tcam) {
+               /* TCAM entries */
+               for (i = 0; i < num; i++) {
+                       if (res[i] < hw->blk[blk].prof.count) {
+                               u16 idx = res[i];
+
+                               ice_free_hw_res(hw, type, 1, &idx);
+                               hw->blk[blk].prof.resource_used_hack[res[i]] =
+                                       false;
+                       }
+               }
+
+       } else {
+               /* Profile IDs */
+               for (i = 0; i < num; i++) {
+                       if (res[i] < hw->blk[blk].es.count) {
+                               u16 idx = res[i];
+
+                               ice_free_hw_res(hw, type, 1, &idx);
+                               hw->blk[blk].es.resource_used_hack[res[i]] =
+                                       false;
+                       }
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_alloc_tcam_ent - allocate hardware tcam entry
+ * @hw: pointer to the hw struct
+ * @blk: the block to allocate the tcam for
+ * @tcam_idx: pointer to variable to receive the tcam entry
+ *
+ * This function allocates a new entry in a Profile ID TCAM for a specific
+ * block.
+ */
+static enum ice_status
+ice_alloc_tcam_ent(struct ice_hw *hw, enum ice_block blk, u16 *tcam_idx)
+{
+       u16 res_type;
+
+       if (!ice_tcam_ent_rsrc_type(blk, &res_type))
+               return ICE_ERR_PARAM;
+
+       return ice_alloc_res_workaround(hw, res_type, 1, tcam_idx);
+}
+
+/**
+ * ice_free_tcam_ent - free hardware tcam entry
+ * @hw: pointer to the hw struct
+ * @blk: the block from which to free the tcam entry
+ * @tcam_idx: the tcam entry to free
+ *
+ * This function frees an entry in a Profile ID TCAM for a specific block.
+ */
+static enum ice_status
+ice_free_tcam_ent(struct ice_hw *hw, enum ice_block blk, u16 tcam_idx)
+{
+       u16 res_type;
+
+       if (!ice_tcam_ent_rsrc_type(blk, &res_type))
+               return ICE_ERR_PARAM;
+
+       return ice_free_res_workaround(hw, res_type, 1, &tcam_idx);
+}
+
+/**
+ * ice_alloc_prof_id - allocate profile id
+ * @hw: pointer to the hw struct
+ * @blk: the block to allocate the profile id for
+ * @prof_id: pointer to variable to receive the profile id
+ *
+ * This function allocates a new profile id, which also corresponds to a Field
+ * Vector (Extraction Sequence) entry.
+ */
+static enum ice_status
+ice_alloc_prof_id(struct ice_hw *hw, enum ice_block blk, u8 *prof_id)
+{
+       enum ice_status status;
+       u16 res_type;
+       u16 get_prof;
+
+       if (!ice_prof_id_rsrc_type(blk, &res_type))
+               return ICE_ERR_PARAM;
+
+       status = ice_alloc_res_workaround(hw, res_type, 1, &get_prof);
+       if (!status)
+               *prof_id = (u8)get_prof;
+
+       return status;
+}
+
+/**
+ * ice_free_prof_id - free profile id
+ * @hw: pointer to the hw struct
+ * @blk: the block from which to free the profile id
+ * @prof_id: the profile id to free
+ *
+ * This function frees a profile id, which also corresponds to a Field Vector.
+ */
+static enum ice_status
+ice_free_prof_id(struct ice_hw *hw, enum ice_block blk, u8 prof_id)
+{
+       u16 tmp_prof_id = (u16)prof_id;
+       u16 res_type;
+
+       if (!ice_prof_id_rsrc_type(blk, &res_type))
+               return ICE_ERR_PARAM;
+
+       return ice_free_res_workaround(hw, res_type, 1, &tmp_prof_id);
+       /* The following code is a WORKAROUND until DCR 076 is available.
+        * DCR 076 - Update to Profile ID TCAM Resource Allocation
+        *
+        * Once the DCR 076 changes are available in FW, this code can be
+        * restored. Original code:
+        *
+        * return ice_free_res(hw, res_type, 1, &tmp_prof_id);
+        */
+}
+
+/**
+ * ice_prof_inc_ref - increment reference count for profile
+ * @hw: pointer to the hw struct
+ * @blk: the block from which to free the profile id
+ * @prof_id: the profile id for which to increment the reference count
+ */
+static enum ice_status
+ice_prof_inc_ref(struct ice_hw *hw, enum ice_block blk, u8 prof_id)
+{
+       if (prof_id > hw->blk[blk].es.count)
+               return ICE_ERR_PARAM;
+
+       hw->blk[blk].es.ref_count[prof_id]++;
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_prof_dec_ref - decrement reference count for profile
+ * @hw: pointer to the hw struct
+ * @blk: the block from which to free the profile id
+ * @prof_id: the profile id for which to decrement the reference count
+ */
+static enum ice_status
+ice_prof_dec_ref(struct ice_hw *hw, enum ice_block blk, u8 prof_id)
+{
+       if (prof_id > hw->blk[blk].es.count)
+               return ICE_ERR_PARAM;
+
+       if (hw->blk[blk].es.ref_count[prof_id] > 0) {
+               if (!--hw->blk[blk].es.ref_count[prof_id])
+                       return ice_free_prof_id(hw, blk, prof_id);
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_write_es - write an extraction sequence to hardware
+ * @hw: pointer to the hw struct
+ * @blk: the block in which to write the extraction sequence
+ * @prof_id: the profile id to write
+ * @fv: pointer to the extraction sequence to write
+ */
+static void
+ice_write_es(struct ice_hw *hw, enum ice_block blk, u8 prof_id,
+            struct ice_fv_word *fv)
+{
+       u16 off;
+
+       off = prof_id * hw->blk[blk].es.fvw;
+       ice_memcpy(&hw->blk[blk].es.t[off], fv, hw->blk[blk].es.fvw * 2,
+                  ICE_NONDMA_TO_NONDMA);
+}
+
+/* Block / table section IDs */
+static const u32 ice_blk_sids[ICE_BLK_COUNT][ICE_SID_OFF_COUNT] = {
+       /* SWITCH */
+       {       ICE_SID_XLT1_SW,
+               ICE_SID_XLT2_SW,
+               ICE_SID_PROFID_TCAM_SW,
+               ICE_SID_PROFID_REDIR_SW,
+               ICE_SID_FLD_VEC_SW
+       },
+
+       /* ACL */
+       {       ICE_SID_XLT1_ACL,
+               ICE_SID_XLT2_ACL,
+               ICE_SID_PROFID_TCAM_ACL,
+               ICE_SID_PROFID_REDIR_ACL,
+               ICE_SID_FLD_VEC_ACL
+       },
+
+       /* FD */
+       {       ICE_SID_XLT1_FD,
+               ICE_SID_XLT2_FD,
+               ICE_SID_PROFID_TCAM_FD,
+               ICE_SID_PROFID_REDIR_FD,
+               ICE_SID_FLD_VEC_FD
+       },
+
+       /* RSS */
+       {       ICE_SID_XLT1_RSS,
+               ICE_SID_XLT2_RSS,
+               ICE_SID_PROFID_TCAM_RSS,
+               ICE_SID_PROFID_REDIR_RSS,
+               ICE_SID_FLD_VEC_RSS
+       },
+
+       /* PE */
+       {       ICE_SID_XLT1_PE,
+               ICE_SID_XLT2_PE,
+               ICE_SID_PROFID_TCAM_PE,
+               ICE_SID_PROFID_REDIR_PE,
+               ICE_SID_FLD_VEC_PE
+       }
+};
+
+/**
+ * ice_fill_tbl - Reads content of a single table type into database
+ * @hw: pointer to the hardware structure
+ * @block_id: Block ID of the table to copy
+ * @sid: Section ID of the table to copy
+ *
+ * Will attempt to read the entire content of a given table of a single block
+ * into the driver database. We assume that the buffer will always
+ * be as large or larger than the data contained in the package. If
+ * this condition is not met, there is most likely an error in the package
+ * contents.
+ */
+static void ice_fill_tbl(struct ice_hw *hw, enum ice_block block_id, u32 sid)
+{
+       u32 dst_len, sect_len, offset = 0;
+       struct ice_prof_redir_section *pr;
+       struct ice_prof_id_section *pid;
+       struct ice_xlt1_section *xlt1;
+       struct ice_xlt2_section *xlt2;
+       struct ice_sw_fv_section *es;
+       struct ice_pkg_enum state;
+       u8 *src, *dst;
+       void *sect;
+
+       /* if the hw segment pointer is null then the first iteration of
+        * ice_pkg_enum_section() will fail. In this case the Hw tables will
+        * not be filled and return success.
+        */
+       if (!hw->seg) {
+               ice_debug(hw, ICE_DBG_PKG, "hw->seg is NULL, tables are not filled\n");
+               return;
+       }
+
+       ice_memset(&state, 0, sizeof(state), ICE_NONDMA_MEM);
+
+       sect = ice_pkg_enum_section(hw->seg, &state, sid);
+
+       while (sect) {
+               switch (sid) {
+               case ICE_SID_XLT1_FD:
+               case ICE_SID_XLT1_RSS:
+               case ICE_SID_XLT1_ACL:
+                       xlt1 = (struct ice_xlt1_section *)sect;
+                       src = xlt1->value;
+                       sect_len = LE16_TO_CPU(xlt1->count) *
+                               sizeof(*hw->blk[block_id].xlt1.t);
+                       dst = hw->blk[block_id].xlt1.t;
+                       dst_len = hw->blk[block_id].xlt1.count *
+                               sizeof(*hw->blk[block_id].xlt1.t);
+                       break;
+               case ICE_SID_XLT2_FD:
+               case ICE_SID_XLT2_RSS:
+               case ICE_SID_XLT2_ACL:
+                       xlt2 = (struct ice_xlt2_section *)sect;
+                       src = (u8 *)xlt2->value;
+                       sect_len = LE16_TO_CPU(xlt2->count) *
+                               sizeof(*hw->blk[block_id].xlt2.t);
+                       dst = (u8 *)hw->blk[block_id].xlt2.t;
+                       dst_len = hw->blk[block_id].xlt2.count *
+                               sizeof(*hw->blk[block_id].xlt2.t);
+                       break;
+               case ICE_SID_PROFID_TCAM_FD:
+               case ICE_SID_PROFID_TCAM_RSS:
+               case ICE_SID_PROFID_TCAM_ACL:
+                       pid = (struct ice_prof_id_section *)sect;
+                       src = (u8 *)pid->entry;
+                       sect_len = LE16_TO_CPU(pid->count) *
+                               sizeof(*hw->blk[block_id].prof.t);
+                       dst = (u8 *)hw->blk[block_id].prof.t;
+                       dst_len = hw->blk[block_id].prof.count *
+                               sizeof(*hw->blk[block_id].prof.t);
+                       break;
+               case ICE_SID_PROFID_REDIR_FD:
+               case ICE_SID_PROFID_REDIR_RSS:
+               case ICE_SID_PROFID_REDIR_ACL:
+                       pr = (struct ice_prof_redir_section *)sect;
+                       src = pr->redir_value;
+                       sect_len = LE16_TO_CPU(pr->count) *
+                               sizeof(*hw->blk[block_id].prof_redir.t);
+                       dst = hw->blk[block_id].prof_redir.t;
+                       dst_len = hw->blk[block_id].prof_redir.count *
+                               sizeof(*hw->blk[block_id].prof_redir.t);
+                       break;
+               case ICE_SID_FLD_VEC_FD:
+               case ICE_SID_FLD_VEC_RSS:
+               case ICE_SID_FLD_VEC_ACL:
+                       es = (struct ice_sw_fv_section *)sect;
+                       src = (u8 *)es->fv;
+                       sect_len = LE16_TO_CPU(es->count) *
+                               sizeof(*hw->blk[block_id].prof_redir.t);
+                       dst = (u8 *)hw->blk[block_id].es.t;
+                       dst_len = hw->blk[block_id].es.count *
+                               sizeof(*hw->blk[block_id].es.t);
+                       break;
+               default:
+                       return;
+               }
+
+               /* if the section offset exceeds destination length, terminate
+                * table fill.
+                */
+               if (offset > dst_len)
+                       return;
+
+               /* if the sum of section size and offset exceed destination size
+                * then we are out of bounds of the Hw table size for that PF.
+                * Changing section length to fill the remaining table space
+                * of that PF.
+                */
+               if ((offset + sect_len) > dst_len)
+                       sect_len = dst_len - offset;
+
+               ice_memcpy(dst + offset, src, sect_len, ICE_NONDMA_TO_NONDMA);
+               offset += sect_len;
+               sect = ice_pkg_enum_section(NULL, &state, sid);
+       }
+}
+
+/**
+ * ice_fill_blk_tbls - Read package content for tables of a block
+ * @hw: pointer to the hardware structure
+ * @block_id: The block ID which contains the tables to be copied
+ *
+ * Reads the current package contents and populates the driver
+ * database with the data it contains to allow for advanced driver
+ * features.
+ */
+static void ice_fill_blk_tbls(struct ice_hw *hw, enum ice_block block_id)
+{
+       ice_fill_tbl(hw, block_id, hw->blk[block_id].xlt1.sid);
+       ice_fill_tbl(hw, block_id, hw->blk[block_id].xlt2.sid);
+       ice_fill_tbl(hw, block_id, hw->blk[block_id].prof.sid);
+       ice_fill_tbl(hw, block_id, hw->blk[block_id].prof_redir.sid);
+       ice_fill_tbl(hw, block_id, hw->blk[block_id].es.sid);
+}
+
+/**
+ * ice_free_prof_map - frees the profile map
+ * @hw: pointer to the hardware structure
+ * @blk: the hw block which contains the profile map to be freed
+ */
+static void ice_free_prof_map(struct ice_hw *hw, enum ice_block blk)
+{
+       struct ice_prof_map *del, *tmp;
+
+       if (LIST_EMPTY(&hw->blk[blk].es.prof_map))
+               return;
+
+       LIST_FOR_EACH_ENTRY_SAFE(del, tmp, &hw->blk[blk].es.prof_map,
+                                ice_prof_map, list) {
+               LIST_DEL(&del->list);
+               ice_free(hw, del);
+       }
+}
+
+/**
+ * ice_free_vsig_tbl - free complete VSIG table entries
+ * @hw: pointer to the hardware structure
+ * @blk: the hw block on which to free the VSIG table entries
+ */
+static void ice_free_vsig_tbl(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 i;
+
+       if (!hw->blk[blk].xlt2.vsig_tbl)
+               return;
+
+       for (i = 1; i < ICE_MAX_VSIGS; i++)
+               if (hw->blk[blk].xlt2.vsig_tbl[i].in_use)
+                       ice_vsig_free(hw, blk, i);
+}
+
+/**
+ * ice_free_hw_tbls - free hardware table memory
+ * @hw: pointer to the hardware structure
+ */
+void ice_free_hw_tbls(struct ice_hw *hw)
+{
+       u8 i;
+
+       for (i = 0; i < ICE_BLK_COUNT; i++) {
+               ice_free_prof_map(hw, (enum ice_block)i);
+               ice_free_vsig_tbl(hw, (enum ice_block)i);
+               ice_free(hw, hw->blk[i].xlt1.ptypes);
+               ice_free(hw, hw->blk[i].xlt1.ptg_tbl);
+               ice_free(hw, hw->blk[i].xlt1.t);
+               ice_free(hw, hw->blk[i].xlt2.t);
+               ice_free(hw, hw->blk[i].xlt2.vsig_tbl);
+               ice_free(hw, hw->blk[i].xlt2.vsis);
+               ice_free(hw, hw->blk[i].prof.t);
+               ice_free(hw, hw->blk[i].prof_redir.t);
+               ice_free(hw, hw->blk[i].es.t);
+               ice_free(hw, hw->blk[i].es.ref_count);
+
+               ice_free(hw, hw->blk[i].es.resource_used_hack);
+               ice_free(hw, hw->blk[i].prof.resource_used_hack);
+       }
+
+       ice_memset(hw->blk, 0, sizeof(hw->blk), ICE_NONDMA_MEM);
+}
+
+/**
+ * ice_init_flow_profs - init flow profile locks and list heads
+ * @hw: pointer to the hardware structure
+ */
+static void ice_init_flow_profs(struct ice_hw *hw)
+{
+       u8 i;
+
+       for (i = 0; i < ICE_BLK_COUNT; i++) {
+               ice_init_lock(&hw->fl_profs_locks[i]);
+               INIT_LIST_HEAD(&hw->fl_profs[i]);
+       }
+}
+
+/**
+ * ice_init_sw_xlt1_db - init software xlt1 database from hw tables
+ * @hw: pointer to the hardware structure
+ * @blk: the hw block to initialize
+ */
+static
+void ice_init_sw_xlt1_db(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 pt;
+
+       for (pt = 0; pt < hw->blk[blk].xlt1.count; pt++) {
+               u8 ptg;
+
+               ptg = hw->blk[blk].xlt1.t[pt];
+               if (ptg != ICE_DEFAULT_PTG) {
+                       ice_ptg_alloc_val(hw, blk, ptg);
+                       ice_ptg_add_mv_ptype(hw, blk, pt, ptg);
+               }
+       }
+}
+
+/**
+ * ice_init_sw_xlt2_db - init software xlt2 database from hw tables
+ * @hw: pointer to the hardware structure
+ * @blk: the hw block to initialize
+ */
+static
+void ice_init_sw_xlt2_db(struct ice_hw *hw, enum ice_block blk)
+{
+       u16 vsi;
+
+       for (vsi = 0; vsi < hw->blk[blk].xlt2.count; vsi++) {
+               u16 vsig;
+
+               vsig = hw->blk[blk].xlt2.t[vsi];
+               if (vsig) {
+                       ice_vsig_alloc_val(hw, blk, vsig);
+                       ice_vsig_add_mv_vsi(hw, blk, vsi, vsig);
+                       /* no changes at this time, since this has been
+                        * initialized from the original package
+                        */
+                       hw->blk[blk].xlt2.vsis[vsi].changed = 0;
+               }
+       }
+}
+
+/**
+ * ice_init_sw_db - init software database from hw tables
+ * @hw: pointer to the hardware structure
+ */
+static
+void ice_init_sw_db(struct ice_hw *hw)
+{
+       u16 i;
+
+       for (i = 0; i < ICE_BLK_COUNT; i++) {
+               ice_init_sw_xlt1_db(hw, (enum ice_block)i);
+               ice_init_sw_xlt2_db(hw, (enum ice_block)i);
+       }
+}
+
+/**
+ * ice_init_hw_tbls - init hardware table memory
+ * @hw: pointer to the hardware structure
+ */
+enum ice_status ice_init_hw_tbls(struct ice_hw *hw)
+{
+       u8 i;
+
+       ice_init_flow_profs(hw);
+
+       for (i = 0; i < ICE_BLK_COUNT; i++) {
+               struct ice_prof_redir *prof_redir = &hw->blk[i].prof_redir;
+               struct ice_prof_tcam *prof = &hw->blk[i].prof;
+               struct ice_xlt1 *xlt1 = &hw->blk[i].xlt1;
+               struct ice_xlt2 *xlt2 = &hw->blk[i].xlt2;
+               struct ice_es *es = &hw->blk[i].es;
+
+               hw->blk[i].overwrite = blk_sizes[i].overwrite;
+               es->reverse = blk_sizes[i].reverse;
+
+               xlt1->sid = ice_blk_sids[i][ICE_SID_XLT1_OFF];
+               xlt1->count = blk_sizes[i].xlt1;
+
+               xlt1->ptypes = (struct ice_ptg_ptype *)
+                       ice_calloc(hw, xlt1->count, sizeof(*xlt1->ptypes));
+
+               if (!xlt1->ptypes)
+                       goto err;
+
+               xlt1->ptg_tbl = (struct ice_ptg_entry *)
+                       ice_calloc(hw, ICE_MAX_PTGS, sizeof(*xlt1->ptg_tbl));
+
+               if (!xlt1->ptg_tbl)
+                       goto err;
+
+               xlt1->t = (u8 *)ice_calloc(hw, xlt1->count, sizeof(*xlt1->t));
+               if (!xlt1->t)
+                       goto err;
+
+               xlt2->sid = ice_blk_sids[i][ICE_SID_XLT2_OFF];
+               xlt2->count = blk_sizes[i].xlt2;
+
+               xlt2->vsis = (struct ice_vsig_vsi *)
+                       ice_calloc(hw, xlt2->count, sizeof(*xlt2->vsis));
+
+               if (!xlt2->vsis)
+                       goto err;
+
+               xlt2->vsig_tbl = (struct ice_vsig_entry *)
+                       ice_calloc(hw, xlt2->count, sizeof(*xlt2->vsig_tbl));
+               if (!xlt2->vsig_tbl)
+                       goto err;
+
+               xlt2->t = (u16 *)ice_calloc(hw, xlt2->count, sizeof(*xlt2->t));
+               if (!xlt2->t)
+                       goto err;
+
+               prof->sid = ice_blk_sids[i][ICE_SID_PR_OFF];
+               prof->count = blk_sizes[i].prof_tcam;
+               prof->max_prof_id = blk_sizes[i].prof_id;
+               prof->cdid_bits = blk_sizes[i].prof_cdid_bits;
+               prof->t = (struct ice_prof_tcam_entry *)
+                       ice_calloc(hw, prof->count, sizeof(*prof->t));
+
+               if (!prof->t)
+                       goto err;
+
+               prof_redir->sid = ice_blk_sids[i][ICE_SID_PR_REDIR_OFF];
+               prof_redir->count = blk_sizes[i].prof_redir;
+               prof_redir->t = (u8 *)ice_calloc(hw, prof_redir->count,
+                                                sizeof(*prof_redir->t));
+
+               if (!prof_redir->t)
+                       goto err;
+
+               es->sid = ice_blk_sids[i][ICE_SID_ES_OFF];
+               es->count = blk_sizes[i].es;
+               es->fvw = blk_sizes[i].fvw;
+               es->t = (struct ice_fv_word *)
+                       ice_calloc(hw, es->count * es->fvw, sizeof(*es->t));
+
+               if (!es->t)
+                       goto err;
+
+               es->ref_count = (u16 *)
+                       ice_calloc(hw, es->count, sizeof(*es->ref_count));
+
+               if (!es->ref_count)
+                       goto err;
+
+               es->resource_used_hack = (u8 *)
+                       ice_calloc(hw, hw->blk[i].es.count, sizeof(u8));
+
+               if (!es->resource_used_hack)
+                       goto err;
+
+               prof->resource_used_hack = (u8 *)ice_calloc(hw, prof->count,
+                                                           sizeof(u8));
+
+               if (!prof->resource_used_hack)
+                       goto err;
+
+               INIT_LIST_HEAD(&es->prof_map);
+
+               /* Now that tables are allocated, read in package data */
+               ice_fill_blk_tbls(hw, (enum ice_block)i);
+       }
+
+       ice_init_sw_db(hw);
+
+       return ICE_SUCCESS;
+
+err:
+       ice_free_hw_tbls(hw);
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_prof_gen_key - generate profile id key
+ * @hw: pointer to the hw struct
+ * @blk: the block in which to write profile id to
+ * @ptg: packet type group (PTG) portion of key
+ * @vsig: VSIG portion of key
+ * @cdid: cdid portion of key
+ * @flags: flag portion of key
+ * @vl_msk: valid mask
+ * @dc_msk: don't care mask
+ * @nm_msk: never match mask
+ * @key: output of profile id key
+ */
+static enum ice_status
+ice_prof_gen_key(struct ice_hw *hw, enum ice_block blk, u8 ptg, u16 vsig,
+                u8 cdid, u16 flags, u8 vl_msk[ICE_TCAM_KEY_VAL_SZ],
+                u8 dc_msk[ICE_TCAM_KEY_VAL_SZ], u8 nm_msk[ICE_TCAM_KEY_VAL_SZ],
+                u8 key[ICE_TCAM_KEY_SZ])
+{
+       struct ice_prof_id_key inkey;
+
+       inkey.xlt1 = ptg;
+       inkey.xlt2_cdid = CPU_TO_LE16(vsig);
+       inkey.flags = CPU_TO_LE16(flags);
+
+       switch (hw->blk[blk].prof.cdid_bits) {
+       case 0:
+               break;
+       case 2:
+#define ICE_CD_2_M 0xC000U
+#define ICE_CD_2_S 14
+               inkey.xlt2_cdid &= ~CPU_TO_LE16(ICE_CD_2_M);
+               inkey.xlt2_cdid |= CPU_TO_LE16(BIT(cdid) << ICE_CD_2_S);
+               break;
+       case 4:
+#define ICE_CD_4_M 0xF000U
+#define ICE_CD_4_S 12
+               inkey.xlt2_cdid &= ~CPU_TO_LE16(ICE_CD_4_M);
+               inkey.xlt2_cdid |= CPU_TO_LE16(BIT(cdid) << ICE_CD_4_S);
+               break;
+       case 8:
+#define ICE_CD_8_M 0xFF00U
+#define ICE_CD_8_S 16
+               inkey.xlt2_cdid &= ~CPU_TO_LE16(ICE_CD_8_M);
+               inkey.xlt2_cdid |= CPU_TO_LE16(BIT(cdid) << ICE_CD_8_S);
+               break;
+       default:
+               ice_debug(hw, ICE_DBG_PKG, "Error in profile config\n");
+               break;
+       };
+
+       return ice_set_key(key, ICE_TCAM_KEY_SZ, (u8 *)&inkey, vl_msk, dc_msk,
+                          nm_msk, 0, ICE_TCAM_KEY_SZ / 2);
+}
+
+/**
+ * ice_tcam_write_entry - write tcam entry
+ * @hw: pointer to the hw struct
+ * @blk: the block in which to write profile id to
+ * @idx: the entry index to write to
+ * @prof_id: profile id
+ * @ptg: packet type group (PTG) portion of key
+ * @vsig: VSIG portion of key
+ * @cdid: cdid portion of key
+ * @flags: flag portion of key
+ * @vl_msk: valid mask
+ * @dc_msk: don't care mask
+ * @nm_msk: never match mask
+ */
+static enum ice_status
+ice_tcam_write_entry(struct ice_hw *hw, enum ice_block blk, u16 idx,
+                    u8 prof_id, u8 ptg, u16 vsig, u8 cdid, u16 flags,
+                    u8 vl_msk[ICE_TCAM_KEY_VAL_SZ],
+                    u8 dc_msk[ICE_TCAM_KEY_VAL_SZ],
+                    u8 nm_msk[ICE_TCAM_KEY_VAL_SZ])
+{
+       struct ice_prof_tcam_entry;
+       enum ice_status status;
+
+       status = ice_prof_gen_key(hw, blk, ptg, vsig, cdid, flags, vl_msk,
+                                 dc_msk, nm_msk, hw->blk[blk].prof.t[idx].key);
+       if (!status) {
+               hw->blk[blk].prof.t[idx].addr = CPU_TO_LE16(idx);
+               hw->blk[blk].prof.t[idx].prof_id = prof_id;
+       }
+
+       return status;
+}
+
+/**
+ * ice_vsig_get_ref - returns number of vsis belong to a VSIG
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsig: VSIG to query
+ * @refs: pointer to variable to receive the reference count
+ */
+static enum ice_status
+ice_vsig_get_ref(struct ice_hw *hw, enum ice_block blk, u16 vsig, u16 *refs)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+       struct ice_vsig_vsi *ptr;
+       *refs = 0;
+
+       if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       ptr = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi;
+       while (ptr) {
+               (*refs)++;
+               ptr = ptr->next_vsi;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_get_ptg - get or allocate a ptg for a ptype
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @ptype: the ptype to retrieve the PTG for
+ * @ptg: receives the PTG of the ptype
+ * @add: receive boolean indicating whether PTG was added or not
+ */
+static enum ice_status
+ice_get_ptg(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 *ptg,
+           bool *add)
+{
+       enum ice_status status;
+
+       *ptg = ICE_DEFAULT_PTG;
+       *add = false;
+
+       status = ice_ptg_find_ptype(hw, blk, ptype, ptg);
+       if (status)
+               return status;
+
+       if (*ptg == ICE_DEFAULT_PTG) {
+               /* need to allocate a PTG, and add ptype to it */
+               *ptg = ice_ptg_alloc(hw, blk);
+               if (*ptg == ICE_DEFAULT_PTG)
+                       return ICE_ERR_HW_TABLE;
+
+               status = ice_ptg_add_mv_ptype(hw, blk, ptype, *ptg);
+               if (status)
+                       return ICE_ERR_HW_TABLE;
+
+               *add = true;
+       }
+
+       return ICE_SUCCESS;
+};
+
+/**
+ * ice_has_prof_vsig - check to see if VSIG has a specific profile
+ * @hw: pointer to the hardware structure
+ * @blk: hw block
+ * @vsig: VSIG to check against
+ * @hdl: profile handle
+ */
+static bool
+ice_has_prof_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+       struct ice_vsig_prof *ent;
+
+       LIST_FOR_EACH_ENTRY(ent, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                           ice_vsig_prof, list) {
+               if (ent->profile_cookie == hdl)
+                       return true;
+       }
+
+       ice_debug(hw, ICE_DBG_INIT,
+                 "Characteristic list for vsi group %d not found.\n",
+                 vsig);
+       return false;
+}
+
+/**
+ * ice_prof_bld_es - build profile id extraction sequence changes
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @bld: the update package buffer build to add to
+ * @chgs: the list of changes to make in hardware
+ */
+static enum ice_status
+ice_prof_bld_es(struct ice_hw *hw, enum ice_block blk,
+               struct ice_buf_build *bld, struct LIST_HEAD_TYPE *chgs)
+{
+       u16 vec_size = hw->blk[blk].es.fvw * sizeof(struct ice_fv_word);
+       struct ice_chs_chg *tmp;
+
+       LIST_FOR_EACH_ENTRY(tmp, chgs, ice_chs_chg, list_entry) {
+               if (tmp->type == ICE_PTG_ES_ADD && tmp->add_prof) {
+                       u16 off = tmp->prof_id * hw->blk[blk].es.fvw;
+                       struct ice_pkg_es *p;
+                       u32 id;
+
+                       id = ice_sect_id(blk, ICE_VEC_TBL);
+                       p = (struct ice_pkg_es *)
+                               ice_pkg_buf_alloc_section(bld, id, sizeof(*p) +
+                                                         vec_size -
+                                                         sizeof(p->es[0]));
+
+                       if (!p)
+                               return ICE_ERR_MAX_LIMIT;
+
+                       p->count = CPU_TO_LE16(1);
+                       p->offset = CPU_TO_LE16(tmp->prof_id);
+
+                       ice_memcpy(p->es, &hw->blk[blk].es.t[off], vec_size,
+                                  ICE_NONDMA_TO_NONDMA);
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_prof_bld_tcam - build profile id tcam changes
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @bld: the update package buffer build to add to
+ * @chgs: the list of changes to make in hardware
+ */
+static enum ice_status
+ice_prof_bld_tcam(struct ice_hw *hw, enum ice_block blk,
+                 struct ice_buf_build *bld, struct LIST_HEAD_TYPE *chgs)
+{
+       struct ice_chs_chg *tmp;
+
+       LIST_FOR_EACH_ENTRY(tmp, chgs, ice_chs_chg, list_entry) {
+               if ((tmp->type == ICE_TCAM_ADD && tmp->add_tcam_idx) ||
+                   tmp->type == ICE_TCAM_REM) {
+                       struct ice_prof_id_section *p;
+                       u32 id;
+
+                       id = ice_sect_id(blk, ICE_PROF_TCAM);
+                       p = (struct ice_prof_id_section *)
+                               ice_pkg_buf_alloc_section(bld, id, sizeof(*p));
+
+                       if (!p)
+                               return ICE_ERR_MAX_LIMIT;
+
+                       p->count = CPU_TO_LE16(1);
+                       p->entry[0].addr = CPU_TO_LE16(tmp->tcam_idx);
+                       p->entry[0].prof_id = tmp->prof_id;
+
+                       ice_memcpy(p->entry[0].key,
+                                  &hw->blk[blk].prof.t[tmp->tcam_idx].key,
+                                  sizeof(hw->blk[blk].prof.t->key),
+                                  ICE_NONDMA_TO_NONDMA);
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_prof_bld_xlt1 - build xlt1 changes
+ * @blk: hardware block
+ * @bld: the update package buffer build to add to
+ * @chgs: the list of changes to make in hardware
+ */
+static enum ice_status
+ice_prof_bld_xlt1(enum ice_block blk, struct ice_buf_build *bld,
+                 struct LIST_HEAD_TYPE *chgs)
+{
+       struct ice_chs_chg *tmp;
+
+       LIST_FOR_EACH_ENTRY(tmp, chgs, ice_chs_chg, list_entry) {
+               if (tmp->type == ICE_PTG_ES_ADD && tmp->add_ptg) {
+                       struct ice_xlt1_section *p;
+                       u32 id;
+
+                       id = ice_sect_id(blk, ICE_XLT1);
+                       p = (struct ice_xlt1_section *)
+                               ice_pkg_buf_alloc_section(bld, id, sizeof(*p));
+
+                       if (!p)
+                               return ICE_ERR_MAX_LIMIT;
+
+                       p->count = CPU_TO_LE16(1);
+                       p->offset = CPU_TO_LE16(tmp->ptype);
+                       p->value[0] = tmp->ptg;
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_prof_bld_xlt2 - build xlt2 changes
+ * @blk: hardware block
+ * @bld: the update package buffer build to add to
+ * @chgs: the list of changes to make in hardware
+ */
+static enum ice_status
+ice_prof_bld_xlt2(enum ice_block blk, struct ice_buf_build *bld,
+                 struct LIST_HEAD_TYPE *chgs)
+{
+       struct ice_chs_chg *tmp;
+
+       LIST_FOR_EACH_ENTRY(tmp, chgs, ice_chs_chg, list_entry) {
+               bool found = false;
+
+               if (tmp->type == ICE_VSIG_ADD)
+                       found = true;
+               else if (tmp->type == ICE_VSI_MOVE)
+                       found = true;
+               else if (tmp->type == ICE_VSIG_REM)
+                       found = true;
+
+               if (found) {
+                       struct ice_xlt2_section *p;
+                       u32 id;
+
+                       id = ice_sect_id(blk, ICE_XLT2);
+                       p = (struct ice_xlt2_section *)
+                               ice_pkg_buf_alloc_section(bld, id, sizeof(*p));
+
+                       if (!p)
+                               return ICE_ERR_MAX_LIMIT;
+
+                       p->count = CPU_TO_LE16(1);
+                       p->offset = CPU_TO_LE16(tmp->vsi);
+                       p->value[0] = CPU_TO_LE16(tmp->vsig);
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_upd_prof_hw - update hardware using the change list
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @chgs: the list of changes to make in hardware
+ */
+static enum ice_status
+ice_upd_prof_hw(struct ice_hw *hw, enum ice_block blk,
+               struct LIST_HEAD_TYPE *chgs)
+{
+       struct ice_buf_build *b;
+       struct ice_chs_chg *tmp;
+       enum ice_status status;
+       u16 pkg_sects = 0;
+       u16 sects = 0;
+       u16 xlt1 = 0;
+       u16 xlt2 = 0;
+       u16 tcam = 0;
+       u16 es = 0;
+
+       /* count number of sections we need */
+       LIST_FOR_EACH_ENTRY(tmp, chgs, ice_chs_chg, list_entry) {
+               switch (tmp->type) {
+               case ICE_PTG_ES_ADD:
+                       if (tmp->add_ptg)
+                               xlt1++;
+                       if (tmp->add_prof)
+                               es++;
+                       break;
+               case ICE_TCAM_ADD:
+               case ICE_TCAM_REM:
+                       tcam++;
+                       break;
+               case ICE_VSIG_ADD:
+               case ICE_VSI_MOVE:
+               case ICE_VSIG_REM:
+                       xlt2++;
+                       break;
+               default:
+                       break;
+               }
+       }
+       sects = xlt1 + xlt2 + tcam + es;
+
+       if (!sects)
+               return ICE_SUCCESS;
+
+       /* Build update package buffer */
+       b = ice_pkg_buf_alloc(hw);
+       if (!b)
+               return ICE_ERR_NO_MEMORY;
+
+       status = ice_pkg_buf_reserve_section(b, sects);
+       if (status)
+               goto error_tmp;
+
+       /* Preserve order of table update: ES, TCAM, PTG, VSIG */
+       if (es) {
+               status = ice_prof_bld_es(hw, blk, b, chgs);
+               if (status)
+                       goto error_tmp;
+       }
+
+       if (tcam) {
+               status = ice_prof_bld_tcam(hw, blk, b, chgs);
+               if (status)
+                       goto error_tmp;
+       }
+
+       if (xlt1) {
+               status = ice_prof_bld_xlt1(blk, b, chgs);
+               if (status)
+                       goto error_tmp;
+       }
+
+       if (xlt2) {
+               status = ice_prof_bld_xlt2(blk, b, chgs);
+               if (status)
+                       goto error_tmp;
+       }
+
+       /* After package buffer build check if the section count in buffer is
+        * non-zero and matches the number of sections detected for package
+        * update.
+        */
+       pkg_sects = ice_pkg_buf_get_active_sections(b);
+       if (!pkg_sects || pkg_sects != sects) {
+               status = ICE_ERR_INVAL_SIZE;
+               goto error_tmp;
+       }
+
+       /* update package */
+       status = ice_update_pkg(hw, ice_pkg_buf(b), 1);
+       if (status == ICE_ERR_AQ_ERROR)
+               ice_debug(hw, ICE_DBG_INIT, "Unable to update HW profile.");
+
+error_tmp:
+       ice_pkg_buf_free(hw, b);
+       return status;
+}
+
+/**
+ * ice_add_prof - add profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking id
+ * @ptypes: array of bitmaps indicating ptypes (ICE_FLOW_PTYPE_MAX bits)
+ * @es: extraction sequence (length of array is determined by the block)
+ *
+ * This function registers a profile, which matches a set of PTYPES with a
+ * particular extraction sequence. While the hardware profile is allocated
+ * it will not be written until the first call to ice_add_flow that specifies
+ * the id value used here.
+ */
+enum ice_status
+ice_add_prof(struct ice_hw *hw, enum ice_block blk, u64 id, u8 ptypes[],
+            struct ice_fv_word *es)
+{
+       u32 bytes = DIVIDE_AND_ROUND_UP(ICE_FLOW_PTYPE_MAX, BITS_PER_BYTE);
+       struct ice_prof_map *prof;
+       enum ice_status status;
+       u32 byte = 0;
+       u8 prof_id;
+
+       /* search for existing profile */
+       status = ice_find_prof_id(hw, blk, es, &prof_id);
+       if (status) {
+               /* allocate profile id */
+               status = ice_alloc_prof_id(hw, blk, &prof_id);
+               if (status)
+                       goto err_ice_add_prof;
+
+               /* and write new es */
+               ice_write_es(hw, blk, prof_id, es);
+       }
+
+       /* add profile info */
+
+       prof = (struct ice_prof_map *)ice_malloc(hw, sizeof(*prof));
+       if (!prof)
+               goto err_ice_add_prof;
+
+       prof->profile_cookie = id;
+       prof->prof_id = prof_id;
+       prof->ptype_count = 0;
+       prof->context = 0;
+
+       /* build list of ptgs */
+       while (bytes && prof->ptype_count < ICE_MAX_PTYPE_PER_PROFILE) {
+               u32 bit;
+
+               if (!ptypes[byte]) {
+                       bytes--;
+                       byte++;
+                       continue;
+               }
+               /* Examine 8 bits per byte */
+               for (bit = 0; bit < 8; bit++) {
+                       if (ptypes[byte] & 1 << bit) {
+                               u16 ptype;
+                               u8 m;
+
+                               ptype = byte * 8 + bit;
+                               if (ptype < ICE_FLOW_PTYPE_MAX) {
+                                       prof->ptype[prof->ptype_count] = ptype;
+
+                                       if (++prof->ptype_count >=
+                                               ICE_MAX_PTYPE_PER_PROFILE)
+                                               break;
+                               }
+
+                               /* nothing left in byte, then exit */
+                               m = ~((1 << (bit + 1)) - 1);
+                               if (!(ptypes[byte] & m))
+                                       break;
+                       }
+               }
+
+               bytes--;
+               byte++;
+       }
+       LIST_ADD(&prof->list, &hw->blk[blk].es.prof_map);
+
+       return ICE_SUCCESS;
+
+err_ice_add_prof:
+       return status;
+}
+
+/**
+ * ice_search_prof_id - Search for a profile tracking ID
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking ID
+ *
+ * This will search for a profile tracking ID which was previously added.
+ */
+struct ice_prof_map *
+ice_search_prof_id(struct ice_hw *hw, enum ice_block blk, u64 id)
+{
+       struct ice_prof_map *entry = NULL;
+       struct ice_prof_map *map;
+
+       LIST_FOR_EACH_ENTRY(map, &hw->blk[blk].es.prof_map, ice_prof_map,
+                           list) {
+               if (map->profile_cookie == id) {
+                       entry = map;
+                       break;
+               }
+       }
+
+       return entry;
+}
+
+/**
+ * ice_set_prof_context - Set context for a given profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking ID
+ * @cntxt: context
+ */
+struct ice_prof_map *
+ice_set_prof_context(struct ice_hw *hw, enum ice_block blk, u64 id, u64 cntxt)
+{
+       struct ice_prof_map *entry;
+
+       entry = ice_search_prof_id(hw, blk, id);
+       if (entry)
+               entry->context = cntxt;
+
+       return entry;
+}
+
+/**
+ * ice_get_prof_context - Get context for a given profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking ID
+ * @cntxt: pointer to variable to receive the context
+ */
+struct ice_prof_map *
+ice_get_prof_context(struct ice_hw *hw, enum ice_block blk, u64 id, u64 *cntxt)
+{
+       struct ice_prof_map *entry;
+
+       entry = ice_search_prof_id(hw, blk, id);
+       if (entry)
+               *cntxt = entry->context;
+
+       return entry;
+}
+
+/**
+ * ice_vsig_prof_id_count - count profiles in a VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: VSIG to remove the profile from
+ */
+static u16
+ice_vsig_prof_id_count(struct ice_hw *hw, enum ice_block blk, u16 vsig)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M, count = 0;
+       struct ice_vsig_prof *p;
+
+       LIST_FOR_EACH_ENTRY(p, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                           ice_vsig_prof, list) {
+               count++;
+       }
+
+       return count;
+}
+
+/**
+ * ice_rel_tcam_idx - release a tcam index
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @idx: the index to release
+ */
+static enum ice_status
+ice_rel_tcam_idx(struct ice_hw *hw, enum ice_block blk, u16 idx)
+{
+       /* Masks to invoke a never match entry */
+       u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF };
+       u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFE, 0xFF, 0xFF, 0xFF, 0xFF };
+       u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 };
+       enum ice_status status;
+
+       /* write the tcam entry */
+       status = ice_tcam_write_entry(hw, blk, idx, 0, 0, 0, 0, 0, vl_msk,
+                                     dc_msk, nm_msk);
+       if (status)
+               return status;
+
+       /* release the tcam entry */
+       status = ice_free_tcam_ent(hw, blk, idx);
+
+       return status;
+}
+
+/**
+ * ice_rem_prof_id - remove one profile from a VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: VSIG to remove the profile from
+ * @prof: pointer to profile structure to remove
+ * @chg: pointer to list to record changes
+ */
+static enum ice_status
+ice_rem_prof_id(struct ice_hw *hw, enum ice_block blk, u16 vsig,
+               struct ice_vsig_prof *prof, struct LIST_HEAD_TYPE *chg)
+{
+       enum ice_status status;
+       struct ice_chs_chg *p;
+       u16 i;
+
+       for (i = 0; i < prof->tcam_count; i++) {
+               p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+               if (!p)
+                       goto err_ice_rem_prof_id;
+
+               p->type = ICE_TCAM_REM;
+               p->vsig = vsig;
+               p->prof_id = prof->tcam[i].prof_id;
+               p->tcam_idx = prof->tcam[i].tcam_idx;
+
+               p->ptg = prof->tcam[i].ptg;
+               prof->tcam[i].in_use = false;
+               p->orig_ent = hw->blk[blk].prof.t[p->tcam_idx];
+               status = ice_rel_tcam_idx(hw, blk, p->tcam_idx);
+               if (!status)
+                       status = ice_prof_dec_ref(hw, blk, p->prof_id);
+
+               LIST_ADD(&p->list_entry, chg);
+
+               if (status)
+                       goto err_ice_rem_prof_id;
+       }
+
+       return ICE_SUCCESS;
+
+err_ice_rem_prof_id:
+       /* caller will clean up the change list */
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_rem_vsig - remove VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: the VSIG to remove
+ * @chg: the change list
+ */
+static enum ice_status
+ice_rem_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig,
+            struct LIST_HEAD_TYPE *chg)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+       struct ice_vsig_vsi *vsi_cur;
+       struct ice_vsig_prof *d, *t;
+       enum ice_status status;
+
+       /* remove TCAM entries */
+       LIST_FOR_EACH_ENTRY_SAFE(d, t,
+                                &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                                ice_vsig_prof, list) {
+               status = ice_rem_prof_id(hw, blk, vsig, d, chg);
+               if (status)
+                       goto err_ice_rem_vsig;
+
+               LIST_DEL(&d->list);
+               ice_free(hw, d);
+       }
+
+       /* Move all VSIS associated with this VSIG to the default VSIG */
+       vsi_cur = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi;
+       if (!vsi_cur)
+               return ICE_ERR_CFG;
+
+       do {
+               struct ice_vsig_vsi *tmp = vsi_cur->next_vsi;
+               struct ice_chs_chg *p;
+
+               p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+               if (!p)
+                       goto err_ice_rem_vsig;
+
+               p->type = ICE_VSIG_REM;
+               p->orig_vsig = vsig;
+               p->vsig = ICE_DEFAULT_VSIG;
+               p->vsi = vsi_cur - hw->blk[blk].xlt2.vsis;
+
+               LIST_ADD(&p->list_entry, chg);
+
+               status = ice_vsig_free(hw, blk, vsig);
+               if (status)
+                       return status;
+
+               vsi_cur = tmp;
+       } while (vsi_cur);
+
+       return ICE_SUCCESS;
+
+err_ice_rem_vsig:
+       /* the caller will free up the change list */
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_rem_prof_id_vsig - remove a specific profile from a VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: VSIG to remove the profile from
+ * @hdl: profile handle indicating which profile to remove
+ * @chg: list to receive a record of changes
+ */
+static enum ice_status
+ice_rem_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl,
+                    struct LIST_HEAD_TYPE *chg)
+{
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+       struct ice_vsig_prof *p, *t;
+       enum ice_status status;
+
+       LIST_FOR_EACH_ENTRY_SAFE(p, t,
+                                &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                                ice_vsig_prof, list) {
+               if (p->profile_cookie == hdl) {
+                       if (ice_vsig_prof_id_count(hw, blk, vsig) == 1)
+                               /* this is the last profile, remove the VSIG */
+                               return ice_rem_vsig(hw, blk, vsig, chg);
+
+                       status = ice_rem_prof_id(hw, blk, vsig, p, chg);
+                       if (!status) {
+                               LIST_DEL(&p->list);
+                               ice_free(hw, p);
+                       }
+                       return status;
+               }
+       }
+
+       return ICE_ERR_DOES_NOT_EXIST;
+}
+
+/**
+ * ice_rem_flow_all - remove all flows with a particular profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking ID
+ */
+static enum ice_status
+ice_rem_flow_all(struct ice_hw *hw, enum ice_block blk, u64 id)
+{
+       struct ice_chs_chg *del, *tmp;
+       struct LIST_HEAD_TYPE chg;
+       enum ice_status status;
+       u16 i;
+
+       INIT_LIST_HEAD(&chg);
+
+       for (i = 1; i < ICE_MAX_VSIGS; i++) {
+               if (hw->blk[blk].xlt2.vsig_tbl[i].in_use) {
+                       if (ice_has_prof_vsig(hw, blk, i, id)) {
+                               status = ice_rem_prof_id_vsig(hw, blk, i, id,
+                                                             &chg);
+                               if (status)
+                                       goto err_ice_rem_flow_all;
+                       }
+               }
+       }
+
+       status = ice_upd_prof_hw(hw, blk, &chg);
+
+err_ice_rem_flow_all:
+       LIST_FOR_EACH_ENTRY_SAFE(del, tmp, &chg, ice_chs_chg, list_entry) {
+               LIST_DEL(&del->list_entry);
+               ice_free(hw, del);
+       }
+
+       return status;
+}
+
+/**
+ * ice_rem_prof - remove profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @id: profile tracking ID
+ *
+ * This will remove the profile specified by the id parameter, which was
+ * previously created through ice_add_prof. If any existing entries
+ * are associated with this profile, they will be removed as well.
+ */
+enum ice_status ice_rem_prof(struct ice_hw *hw, enum ice_block blk, u64 id)
+{
+       enum ice_status status;
+       struct ice_prof_map *pmap;
+
+       pmap = ice_search_prof_id(hw, blk, id);
+       if (!pmap)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       status = ice_free_prof_id(hw, blk, pmap->prof_id);
+
+       if (status)
+               return status;
+
+       /* remove all flows with this profile */
+       status = ice_rem_flow_all(hw, blk, pmap->profile_cookie);
+       if (status)
+               return status;
+       LIST_DEL(&pmap->list);
+       ice_free(hw, pmap);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_get_prof_ptgs - get ptgs for profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @hdl: profile handle
+ * @chg: change list
+ */
+static enum ice_status
+ice_get_prof_ptgs(struct ice_hw *hw, enum ice_block blk, u64 hdl,
+                 struct LIST_HEAD_TYPE *chg)
+{
+       struct ice_prof_map *map;
+       struct ice_chs_chg *p;
+       u16 i;
+
+       /* Get the details on the profile specified by the handle id */
+       map = ice_search_prof_id(hw, blk, hdl);
+       if (!map)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       for (i = 0; i < map->ptype_count; i++) {
+               enum ice_status status;
+               bool add;
+               u8 ptg;
+
+               status = ice_get_ptg(hw, blk, map->ptype[i], &ptg, &add);
+               if (status)
+                       goto err_ice_get_prof_ptgs;
+
+               if (add || !hw->blk[blk].es.ref_count[map->prof_id]) {
+                       /* add PTG to change list */
+                       p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+                       if (!p)
+                               goto err_ice_get_prof_ptgs;
+
+                       p->type = ICE_PTG_ES_ADD;
+                       p->ptype = map->ptype[i];
+                       p->ptg = ptg;
+                       p->add_ptg = add;
+
+                       p->add_prof = !hw->blk[blk].es.ref_count[map->prof_id];
+                       p->prof_id = map->prof_id;
+
+                       LIST_ADD(&p->list_entry, chg);
+               }
+       }
+
+       return ICE_SUCCESS;
+
+err_ice_get_prof_ptgs:
+       /* let caller clean up the change list */
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_get_profs_vsig - get a copy of the list of profiles from a VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: VSIG from which to copy the list
+ * @lst: output list
+ *
+ * This routine makes a copy of the list of profiles in the specified VSIG.
+ */
+static enum ice_status
+ice_get_profs_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig,
+                  struct LIST_HEAD_TYPE *lst)
+{
+       struct ice_vsig_prof *ent1, *ent2;
+       u16 idx = vsig & ICE_VSIG_IDX_M;
+
+       LIST_FOR_EACH_ENTRY(ent1, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                           ice_vsig_prof, list) {
+               struct ice_vsig_prof *p;
+
+               /* copy to the input list */
+               p = (struct ice_vsig_prof *)ice_malloc(hw, sizeof(*p));
+               if (!p)
+                       goto err_ice_get_profs_vsig;
+
+               ice_memcpy(p, ent1, sizeof(*p), ICE_NONDMA_TO_NONDMA);
+
+               LIST_ADD(&p->list, lst);
+       }
+
+       return ICE_SUCCESS;
+
+err_ice_get_profs_vsig:
+       LIST_FOR_EACH_ENTRY_SAFE(ent1, ent2, lst, ice_vsig_prof, list) {
+               LIST_DEL(&ent1->list);
+               ice_free(hw, ent1);
+       }
+
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_add_prof_to_lst - add profile entry to a list
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @lst: the list to be added to
+ * @hdl: profile handle of entry to add
+ */
+static enum ice_status
+ice_add_prof_to_lst(struct ice_hw *hw, enum ice_block blk,
+                   struct LIST_HEAD_TYPE *lst, u64 hdl)
+{
+       struct ice_vsig_prof *p;
+       struct ice_prof_map *map;
+       u16 i;
+
+       map = ice_search_prof_id(hw, blk, hdl);
+       if (!map)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       p = (struct ice_vsig_prof *)ice_malloc(hw, sizeof(*p));
+       if (!p)
+               return ICE_ERR_NO_MEMORY;
+
+       p->profile_cookie = map->profile_cookie;
+       p->prof_id = map->prof_id;
+       p->tcam_count = map->ptype_count;
+
+       for (i = 0; i < map->ptype_count; i++) {
+               enum ice_status status;
+               u8 ptg;
+
+               p->tcam[i].prof_id = map->prof_id;
+               p->tcam[i].tcam_idx = ICE_INVALID_TCAM;
+
+               status = ice_ptg_find_ptype(hw, blk, map->ptype[i], &ptg);
+               if (status) {
+                       ice_free(hw, p);
+                       return status;
+               }
+
+               p->tcam[i].ptg = ptg;
+       }
+
+       LIST_ADD(&p->list, lst);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_move_vsi - move VSI to another VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: the VSI to move
+ * @vsig: the VSIG to move the VSI to
+ * @chg: the change list
+ */
+static enum ice_status
+ice_move_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig,
+            struct LIST_HEAD_TYPE *chg)
+{
+       enum ice_status status;
+       struct ice_chs_chg *p;
+       u16 orig_vsig;
+
+       p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+       if (!p)
+               return ICE_ERR_NO_MEMORY;
+
+       status = ice_vsig_find_vsi(hw, blk, vsi, &orig_vsig);
+       if (!status)
+               status = ice_vsig_add_mv_vsi(hw, blk, vsi, vsig);
+       if (status) {
+               ice_free(hw, p);
+               return status;
+       }
+
+       p->type = ICE_VSI_MOVE;
+       p->vsi = vsi;
+       p->orig_vsig = orig_vsig;
+       p->vsig = vsig;
+
+       LIST_ADD(&p->list_entry, chg);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_prof_tcam_ena_dis - add enable or disable tcam change
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @enable: true to enable, false to disable
+ * @vsig: the vsig of the tcam entry
+ * @tcam: pointer the tcam info structure of the tcam to disable
+ * @chg: the change list
+ *
+ * This function appends an enable or disable tcam entry in the change log
+ */
+static enum ice_status
+ice_prof_tcam_ena_dis(struct ice_hw *hw, enum ice_block blk, bool enable,
+                     u16 vsig, struct ice_tcam_inf *tcam,
+                     struct LIST_HEAD_TYPE *chg)
+{
+       enum ice_status status;
+       struct ice_chs_chg *p;
+
+       /* Default: enable means change the low flag bit to don't care */
+       u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 };
+       u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x00, 0x00, 0x00, 0x00, 0x00 };
+       u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 };
+
+       /* If disabled, change the low flag bit to never match */
+       if (!enable) {
+               dc_msk[0] = 0x00;
+               nm_msk[0] = 0x01;
+       }
+
+       /* add TCAM to change list */
+       p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+       if (!p)
+               return ICE_ERR_NO_MEMORY;
+
+       status = ice_tcam_write_entry(hw, blk, tcam->tcam_idx, tcam->prof_id,
+                                     tcam->ptg, vsig, 0, 0, vl_msk, dc_msk,
+                                     nm_msk);
+       if (status)
+               goto err_ice_prof_tcam_ena_dis;
+
+       tcam->in_use = enable;
+
+       p->type = ICE_TCAM_ADD;
+       p->add_tcam_idx = true;
+       p->prof_id = tcam->prof_id;
+       p->ptg = tcam->ptg;
+       p->vsig = 0;
+       p->tcam_idx = tcam->tcam_idx;
+
+       /* log change */
+       LIST_ADD(&p->list_entry, chg);
+
+       return ICE_SUCCESS;
+
+err_ice_prof_tcam_ena_dis:
+       ice_free(hw, p);
+       return status;
+}
+
+/**
+ * ice_adj_prof_priorities - adjust profile based on priorities
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: the VSIG for which to adjust profile priorities
+ * @chg: the change list
+ */
+static enum ice_status
+ice_adj_prof_priorities(struct ice_hw *hw, enum ice_block blk, u16 vsig,
+                       struct LIST_HEAD_TYPE *chg)
+{
+       ice_declare_bitmap(ptgs_used, ICE_XLT1_CNT);
+       struct ice_vsig_prof *t;
+       enum ice_status status;
+       u16 idx;
+
+       ice_memset(ptgs_used, 0, sizeof(ptgs_used), ICE_NONDMA_MEM);
+       idx = vsig & ICE_VSIG_IDX_M;
+
+       /* Priority is based on the order in which the profiles are added. The
+        * newest added profile has highest priority and the oldest added
+        * profile has the lowest priority. Since the profile property list for
+        * a VSIG is sorted from newest to oldest, this code traverses the list
+        * in order and enables the first of each PTG that it finds (that is not
+        * already enabled); it also disables any duplicate PTGs that it finds
+        * in the older profiles (that are currently enabled).
+        */
+
+       LIST_FOR_EACH_ENTRY(t, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst,
+                           ice_vsig_prof, list) {
+               u16 i;
+
+               for (i = 0; i < t->tcam_count; i++) {
+                       /* Scan the priorities from newest to oldest.
+                        * Make sure that the newest profiles take priority.
+                        */
+                       if (ice_is_bit_set(ptgs_used, t->tcam[i].ptg) &&
+                           t->tcam[i].in_use) {
+                               /* need to mark this PTG as never match, as it
+                                * was already in use and therefore duplicate
+                                * (and lower priority)
+                                */
+                               status = ice_prof_tcam_ena_dis(hw, blk, false,
+                                                              vsig,
+                                                              &t->tcam[i],
+                                                              chg);
+                               if (status)
+                                       return status;
+                       } else if (!ice_is_bit_set(ptgs_used, t->tcam[i].ptg) &&
+                                  !t->tcam[i].in_use) {
+                               /* need to enable this PTG, as it in not in use
+                                * and not enabled (highest priority)
+                                */
+                               status = ice_prof_tcam_ena_dis(hw, blk, true,
+                                                              vsig,
+                                                              &t->tcam[i],
+                                                              chg);
+                               if (status)
+                                       return status;
+                       }
+
+                       /* keep track of used ptgs */
+                       ice_set_bit(t->tcam[i].ptg, ptgs_used);
+               }
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_add_prof_id_vsig - add profile to VSIG
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsig: the VSIG to which this profile is to be added
+ * @hdl: the profile handle indicating the profile to add
+ * @chg: the change list
+ */
+static enum ice_status
+ice_add_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl,
+                    struct LIST_HEAD_TYPE *chg)
+{
+       /* Masks that ignore flags */
+       u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF };
+       u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0x00, 0x00, 0x00 };
+       u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x00, 0x00, 0x00, 0x00, 0x00 };
+       struct ice_prof_map *map;
+       struct ice_vsig_prof *t;
+       struct ice_chs_chg *p;
+       u16 i;
+
+       /* Get the details on the profile specified by the handle id */
+       map = ice_search_prof_id(hw, blk, hdl);
+       if (!map)
+               return ICE_ERR_DOES_NOT_EXIST;
+
+       /* Error, if this VSIG already has this profile */
+       if (ice_has_prof_vsig(hw, blk, vsig, hdl))
+               return ICE_ERR_ALREADY_EXISTS;
+
+       /* new VSIG profile structure */
+       t = (struct ice_vsig_prof *)ice_malloc(hw, sizeof(*t));
+       if (!t)
+               goto err_ice_add_prof_id_vsig;
+
+       t->profile_cookie = map->profile_cookie;
+       t->prof_id = map->prof_id;
+       t->tcam_count = map->ptype_count;
+
+       /* create tcam entries */
+       for (i = 0; i < map->ptype_count; i++) {
+               enum ice_status status;
+               u16 tcam_idx;
+               bool add;
+               u8 ptg;
+
+               /* If properly sequenced, we should never have to allocate new
+                * PTGs
+                */
+               status = ice_get_ptg(hw, blk, map->ptype[i], &ptg, &add);
+               if (status)
+                       goto err_ice_add_prof_id_vsig;
+
+               /* add TCAM to change list */
+               p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+               if (!p)
+                       goto err_ice_add_prof_id_vsig;
+
+               /* allocate the tcam entry index */
+               status = ice_alloc_tcam_ent(hw, blk, &tcam_idx);
+               if (status)
+                       goto err_ice_add_prof_id_vsig;
+
+               t->tcam[i].ptg = ptg;
+               t->tcam[i].prof_id = map->prof_id;
+               t->tcam[i].tcam_idx = tcam_idx;
+               t->tcam[i].in_use = true;
+
+               p->type = ICE_TCAM_ADD;
+               p->add_tcam_idx = true;
+               p->prof_id = t->tcam[i].prof_id;
+               p->ptg = t->tcam[i].ptg;
+               p->vsig = vsig;
+               p->tcam_idx = t->tcam[i].tcam_idx;
+
+               /* write the tcam entry */
+               status = ice_tcam_write_entry(hw, blk, t->tcam[i].tcam_idx,
+                                             t->tcam[i].prof_id,
+                                             t->tcam[i].ptg, vsig, 0, 0,
+                                             vl_msk, dc_msk, nm_msk);
+               if (status)
+                       goto err_ice_add_prof_id_vsig;
+
+               /* this increments the reference count of how many tcam entries
+                * are using this hw profile id
+                */
+               status = ice_prof_inc_ref(hw, blk, t->tcam[i].prof_id);
+
+               /* log change */
+               LIST_ADD(&p->list_entry, chg);
+       }
+
+       /* add profile to VSIG */
+       LIST_ADD(&t->list,
+                &hw->blk[blk].xlt2.vsig_tbl[(vsig & ICE_VSIG_IDX_M)].prop_lst);
+
+       return ICE_SUCCESS;
+
+err_ice_add_prof_id_vsig:
+       /* let caller clean up the change list */
+       return ICE_ERR_NO_MEMORY;
+}
+
+/**
+ * ice_create_prof_id_vsig - add a new VSIG with a single profile
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: the initial VSI that will be in VSIG
+ * @hdl: the profile handle of the profile that will be added to the VSIG
+ * @chg: the change list
+ */
+static enum ice_status
+ice_create_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl,
+                       struct LIST_HEAD_TYPE *chg)
+{
+       enum ice_status status;
+       struct ice_chs_chg *p;
+       u16 new_vsig;
+
+       p = (struct ice_chs_chg *)ice_malloc(hw, sizeof(*p));
+       if (!p)
+               return ICE_ERR_NO_MEMORY;
+
+       new_vsig = ice_vsig_alloc(hw, blk);
+       if (!new_vsig)
+               return ICE_ERR_HW_TABLE;
+
+       status = ice_move_vsi(hw, blk, vsi, new_vsig, chg);
+       if (status)
+               return status;
+
+       status = ice_add_prof_id_vsig(hw, blk, new_vsig, hdl, chg);
+       if (status)
+               return status;
+
+       p->type = ICE_VSIG_ADD;
+       p->vsi = vsi;
+       p->orig_vsig = ICE_DEFAULT_VSIG;
+       p->vsig = new_vsig;
+
+       LIST_ADD(&p->list_entry, chg);
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_create_vsig_from_list - create a new VSIG with a list of profiles
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: the initial VSI that will be in VSIG
+ * @lst: the list of profile that will be added to the VSIG
+ * @chg: the change list
+ */
+static enum ice_status
+ice_create_vsig_from_lst(struct ice_hw *hw, enum ice_block blk, u16 vsi,
+                        struct LIST_HEAD_TYPE *lst, struct LIST_HEAD_TYPE *chg)
+{
+       struct ice_vsig_prof *t;
+       enum ice_status status;
+       u16 vsig;
+
+       vsig = ice_vsig_alloc(hw, blk);
+       if (!vsig)
+               return ICE_ERR_HW_TABLE;
+
+       status = ice_move_vsi(hw, blk, vsi, vsig, chg);
+       if (status)
+               return status;
+
+       LIST_FOR_EACH_ENTRY(t, lst, ice_vsig_prof, list) {
+               status = ice_add_prof_id_vsig(hw, blk, vsig, t->profile_cookie,
+                                             chg);
+               if (status)
+                       return status;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_find_prof_vsig - find a VSIG with a specific profile handle
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @hdl: the profile handle of the profile to search for
+ * @vsig: returns the VSIG with the matching profile
+ */
+static bool
+ice_find_prof_vsig(struct ice_hw *hw, enum ice_block blk, u64 hdl, u16 *vsig)
+{
+       struct ice_vsig_prof *t;
+       struct LIST_HEAD_TYPE lst;
+       enum ice_status status;
+
+       INIT_LIST_HEAD(&lst);
+
+       t = (struct ice_vsig_prof *)ice_malloc(hw, sizeof(*t));
+       if (!t)
+               return false;
+
+       t->profile_cookie = hdl;
+       LIST_ADD(&t->list, &lst);
+
+       status = ice_find_dup_props_vsig(hw, blk, &lst, vsig);
+
+       LIST_DEL(&t->list);
+       ice_free(hw, t);
+
+       return status == ICE_SUCCESS;
+}
+
+/**
+ * ice_add_prof_id_flow - add profile flow
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: the vsi to enable with the profile specified by id
+ * @hdl: profile handle
+ *
+ * Calling this function will update the hardware tables to enable the
+ * profile indicated by the id parameter for the VSIs specified in the vsi
+ * array. Once successfully called, the flow will be enabled.
+ */
+enum ice_status
+ice_add_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl)
+{
+       struct ice_vsig_prof *tmp1, *del1;
+       struct LIST_HEAD_TYPE union_lst;
+       struct ice_chs_chg *tmp, *del;
+       struct LIST_HEAD_TYPE chrs;
+       struct LIST_HEAD_TYPE chg;
+       enum ice_status status;
+       u16 vsig, or_vsig = 0;
+
+       INIT_LIST_HEAD(&union_lst);
+       INIT_LIST_HEAD(&chrs);
+       INIT_LIST_HEAD(&chg);
+
+       status = ice_get_prof_ptgs(hw, blk, hdl, &chg);
+       if (status)
+               return status;
+
+       /* determine if vsi is already part of a VSIG */
+       status = ice_vsig_find_vsi(hw, blk, vsi, &vsig);
+       if (!status && vsig) {
+               bool only_vsi;
+               u16 ref;
+
+               /* found in vsig */
+               or_vsig = vsig;
+
+               /* make sure that there is no overlap/conflict between the new
+                * characteristics and the existing ones; we don't support that
+                * scenario
+                */
+               if (ice_has_prof_vsig(hw, blk, vsig, hdl)) {
+                       status = ICE_ERR_ALREADY_EXISTS;
+                       goto err_ice_add_prof_id_flow;
+               }
+
+               /* last VSI in the VSIG? */
+               status = ice_vsig_get_ref(hw, blk, vsig, &ref);
+               if (status)
+                       goto err_ice_add_prof_id_flow;
+               only_vsi = (ref == 1);
+
+               /* create a union of the current profiles and the one being
+                * added
+                */
+               status = ice_get_profs_vsig(hw, blk, vsig, &union_lst);
+               if (status)
+                       goto err_ice_add_prof_id_flow;
+
+               status = ice_add_prof_to_lst(hw, blk, &union_lst, hdl);
+               if (status)
+                       goto err_ice_add_prof_id_flow;
+
+               /* search for an existing VSIG with an exact charc match */
+               status = ice_find_dup_props_vsig(hw, blk, &union_lst, &vsig);
+               if (!status) {
+                       /* found an exact match */
+                       /* move vsi to the VSIG that matches */
+                       status = ice_move_vsi(hw, blk, vsi, vsig, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+
+                       /* remove original VSIG if we just moved the only VSI
+                        * from it
+                        */
+                       if (only_vsi) {
+                               status = ice_rem_vsig(hw, blk, or_vsig, &chg);
+                               if (status)
+                                       goto err_ice_add_prof_id_flow;
+                       }
+               } else if (only_vsi) {
+                       /* If the original VSIG only contains one VSI, then it
+                        * will be the requesting VSI. In this case the VSI is
+                        * not sharing entries and we can simply add the new
+                        * profile to the VSIG.
+                        */
+                       status = ice_add_prof_id_vsig(hw, blk, vsig, hdl, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+
+                       /* Adjust priorities */
+                       status = ice_adj_prof_priorities(hw, blk, vsig, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+               } else {
+                       /* No match, so we need a new VSIG */
+                       status = ice_create_vsig_from_lst(hw, blk, vsi,
+                                                         &union_lst, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+
+                       /* Adjust priorities */
+                       status = ice_adj_prof_priorities(hw, blk, vsig, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+               }
+       } else {
+               /* need to find or add a VSIG */
+               /* search for an exising VSIG with an exact charc match */
+               if (ice_find_prof_vsig(hw, blk, hdl, &vsig)) {
+                       /* found an exact match */
+                       /* add or move vsi to the VSIG that matches */
+                       status = ice_move_vsi(hw, blk, vsi, vsig, &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+               } else {
+                       /* we did not find an exact match */
+                       /* we need to add a VSIG */
+                       status = ice_create_prof_id_vsig(hw, blk, vsi, hdl,
+                                                        &chg);
+                       if (status)
+                               goto err_ice_add_prof_id_flow;
+               }
+       }
+
+       /* update hardware */
+       if (!status)
+               status = ice_upd_prof_hw(hw, blk, &chg);
+
+err_ice_add_prof_id_flow:
+       LIST_FOR_EACH_ENTRY_SAFE(del, tmp, &chg, ice_chs_chg, list_entry) {
+               LIST_DEL(&del->list_entry);
+               ice_free(hw, del);
+       }
+
+       LIST_FOR_EACH_ENTRY_SAFE(del1, tmp1, &union_lst, ice_vsig_prof, list) {
+               LIST_DEL(&del1->list);
+               ice_free(hw, del1);
+       }
+
+       LIST_FOR_EACH_ENTRY_SAFE(del1, tmp1, &chrs, ice_vsig_prof, list) {
+               LIST_DEL(&del1->list);
+               ice_free(hw, del1);
+       }
+
+       return status;
+}
+
+/**
+ * ice_add_flow - add flow
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: array of VSIs to enable with the profile specified by id
+ * @count: number of elements in the vsi array
+ * @id: profile tracking id
+ *
+ * Calling this function will update the hardware tables to enable the
+ * profile indicated by the id parameter for the VSIs specified in the vsi
+ * array. Once successfully called, the flow will be enabled.
+ */
+enum ice_status
+ice_add_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi[], u8 count,
+            u64 id)
+{
+       enum ice_status status;
+       u16 i;
+
+       for (i = 0; i < count; i++) {
+               status = ice_add_prof_id_flow(hw, blk, vsi[i], id);
+               if (status)
+                       return status;
+       }
+
+       return ICE_SUCCESS;
+}
+
+/**
+ * ice_rem_prof_from_list - remove a profile from list
+ * @hw: pointer to the hw struct
+ * @lst: list to remove the profile from
+ * @hdl: the profile handle indicating the profile to remove
+ */
+static enum ice_status
+ice_rem_prof_from_list(struct ice_hw *hw, struct LIST_HEAD_TYPE *lst, u64 hdl)
+{
+       struct ice_vsig_prof *ent, *tmp;
+
+       LIST_FOR_EACH_ENTRY_SAFE(ent, tmp, lst, ice_vsig_prof, list) {
+               if (ent->profile_cookie == hdl) {
+                       LIST_DEL(&ent->list);
+                       ice_free(hw, ent);
+                       return ICE_SUCCESS;
+               }
+       }
+
+       return ICE_ERR_DOES_NOT_EXIST;
+}
+
+/**
+ * ice_rem_prof_id_flow - remove flow
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: the vsi from which to remove the profile specified by id
+ * @hdl: profile tracking handle
+ *
+ * Calling this function will update the hardware tables to remove the
+ * profile indicated by the id parameter for the VSIs specified in the vsi
+ * array. Once successfully called, the flow will be disabled.
+ */
+enum ice_status
+ice_rem_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl)
+{
+       struct ice_vsig_prof *tmp1, *del1;
+       struct LIST_HEAD_TYPE chg, copy;
+       struct ice_chs_chg *tmp, *del;
+       enum ice_status status;
+       u16 vsig;
+
+       INIT_LIST_HEAD(&copy);
+       INIT_LIST_HEAD(&chg);
+
+       /* determine if vsi is already part of a VSIG */
+       status = ice_vsig_find_vsi(hw, blk, vsi, &vsig);
+       if (!status && vsig) {
+               bool last_profile;
+               bool only_vsi;
+               u16 ref;
+
+               /* found in VSIG */
+               last_profile = ice_vsig_prof_id_count(hw, blk, vsig) == 1;
+               status = ice_vsig_get_ref(hw, blk, vsig, &ref);
+               if (status)
+                       goto err_ice_rem_prof_id_flow;
+               only_vsi = (ref == 1);
+
+               if (only_vsi) {
+                       /* If the original VSIG only contains one reference,
+                        * which will be the requesting VSI, then the VSI is not
+                        * sharing entries and we can simply remove the specific
+                        * characteristics from the VSIG.
+                        */
+
+                       if (last_profile) {
+                               /* If there are no profiles left for this VSIG,
+                                * then simply remove the the VSIG.
+                                */
+                               status = ice_rem_vsig(hw, blk, vsig, &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+                       } else {
+                               status = ice_rem_prof_id_vsig(hw, blk, vsig,
+                                                             hdl, &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+
+                               /* Adjust priorities */
+                               status = ice_adj_prof_priorities(hw, blk, vsig,
+                                                                &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+                       }
+
+               } else {
+                       /* Make a copy of the VSIG's list of Profiles */
+                       status = ice_get_profs_vsig(hw, blk, vsig, &copy);
+                       if (status)
+                               goto err_ice_rem_prof_id_flow;
+
+                       /* Remove specified profile entry from the list */
+                       status = ice_rem_prof_from_list(hw, &copy, hdl);
+                       if (status)
+                               goto err_ice_rem_prof_id_flow;
+
+                       if (LIST_EMPTY(&copy)) {
+                               status = ice_move_vsi(hw, blk, vsi,
+                                                     ICE_DEFAULT_VSIG, &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+
+                       } else if (ice_find_dup_props_vsig(hw, blk, &copy,
+                                                          &vsig)) {
+                               /* found an exact match */
+                               /* add or move vsi to the VSIG that matches */
+                               /* Search for a VSIG with a matching profile
+                                * list
+                                */
+
+                               /* Found match, move VSI to the matching VSIG */
+                               status = ice_move_vsi(hw, blk, vsi, vsig, &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+                       } else {
+                               /* since no existing VSIG supports this
+                                * characteristic pattern, we need to create a
+                                * new VSIG and tcam entries
+                                */
+                               status = ice_create_vsig_from_lst(hw, blk, vsi,
+                                                                 &copy, &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+
+                               /* Adjust priorities */
+                               status = ice_adj_prof_priorities(hw, blk, vsig,
+                                                                &chg);
+                               if (status)
+                                       goto err_ice_rem_prof_id_flow;
+                       }
+               }
+       } else {
+               status = ICE_ERR_DOES_NOT_EXIST;
+       }
+
+       /* update hardware tables */
+       if (!status)
+               status = ice_upd_prof_hw(hw, blk, &chg);
+
+err_ice_rem_prof_id_flow:
+       LIST_FOR_EACH_ENTRY_SAFE(del, tmp, &chg, ice_chs_chg, list_entry) {
+               LIST_DEL(&del->list_entry);
+               ice_free(hw, del);
+       }
+
+       LIST_FOR_EACH_ENTRY_SAFE(del1, tmp1, &copy, ice_vsig_prof, list) {
+               LIST_DEL(&del1->list);
+               ice_free(hw, del1);
+       }
+
+       return status;
+}
+
+/**
+ * ice_rem_flow - remove flow
+ * @hw: pointer to the hw struct
+ * @blk: hardware block
+ * @vsi: array of VSIs from which to remove the profile specified by id
+ * @count: number of elements in the vsi array
+ * @id: profile tracking id
+ *
+ * The function will remove flows from the specified VSIs that were enabled
+ * using ice_add_flow. The id value will indicated which profile will be
+ * removed. Once successfully called, the flow will be disabled.
+ */
+enum ice_status
+ice_rem_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi[], u8 count,
+            u64 id)
+{
+       enum ice_status status;
+       u16 i;
+
+       for (i = 0; i < count; i++) {
+               status = ice_rem_prof_id_flow(hw, blk, vsi[i], id);
+               if (status)
+                       return status;
+       }
+
+       return ICE_SUCCESS;
+}
diff --git a/drivers/net/ice/base/ice_flex_pipe.h b/drivers/net/ice/base/ice_flex_pipe.h
new file mode 100644 (file)
index 0000000..81e89ae
--- /dev/null
@@ -0,0 +1,107 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2019
+ */
+
+#ifndef _ICE_FLEX_PIPE_H_
+#define _ICE_FLEX_PIPE_H_
+
+#include "ice_type.h"
+
+/* Package format version */
+#define ICE_PKG_FMT_VER_MAJ    1
+#define ICE_PKG_FMT_VER_MNR    0
+#define ICE_PKG_FMT_VER_UPD    0
+#define ICE_PKG_FMT_VER_DFT    0
+
+enum ice_status
+ice_update_pkg(struct ice_hw *hw, struct ice_buf *bufs, u32 count);
+
+struct ice_generic_seg_hdr *
+ice_find_seg_in_pkg(struct ice_hw *hw, u32 seg_type,
+                   struct ice_pkg_hdr *pkg_hdr);
+enum ice_status ice_download_pkg(struct ice_hw *hw, struct ice_seg *ice_seg);
+
+enum ice_status
+ice_init_pkg_info(struct ice_hw *hw, struct ice_pkg_hdr *pkg_header);
+
+void ice_init_pkg_hints(struct ice_hw *hw, struct ice_seg *ice_seg);
+
+enum ice_status
+ice_find_label_value(struct ice_seg *ice_seg, char const *name, u32 type,
+                    u16 *value);
+enum ice_status
+ice_get_sw_fv_list(struct ice_hw *hw, u16 *prot_ids, u8 ids_cnt,
+                  struct LIST_HEAD_TYPE *fv_list);
+enum ice_status
+ice_aq_upload_section(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf,
+                     u16 buf_size, struct ice_sq_cd *cd);
+
+enum ice_status
+ice_pkg_buf_unreserve_section(struct ice_buf_build *bld, u16 count);
+u16 ice_pkg_buf_get_free_space(struct ice_buf_build *bld);
+u16 ice_pkg_buf_get_active_sections(struct ice_buf_build *bld);
+
+/* package buffer building routines */
+
+struct ice_buf_build *ice_pkg_buf_alloc(struct ice_hw *hw);
+enum ice_status
+ice_pkg_buf_reserve_section(struct ice_buf_build *bld, u16 count);
+void *ice_pkg_buf_alloc_section(struct ice_buf_build *bld, u32 type, u16 size);
+struct ice_buf *ice_pkg_buf(struct ice_buf_build *bld);
+void ice_pkg_buf_free(struct ice_hw *hw, struct ice_buf_build *bld);
+
+/* XLT1/PType group functions */
+enum ice_status ice_ptg_update_xlt1(struct ice_hw *hw, enum ice_block blk);
+enum ice_status
+ice_ptg_find_ptype(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 *ptg);
+u8 ice_ptg_alloc(struct ice_hw *hw, enum ice_block blk);
+void ice_ptg_free(struct ice_hw *hw, enum ice_block blk, u8 ptg);
+enum ice_status
+ice_ptg_add_mv_ptype(struct ice_hw *hw, enum ice_block blk, u16 ptype, u8 ptg);
+
+/* XLT2/Vsi group functions */
+enum ice_status ice_vsig_update_xlt2(struct ice_hw *hw, enum ice_block blk);
+enum ice_status
+ice_vsig_find_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 *vsig);
+enum ice_status
+ice_find_dup_props_vsig(struct ice_hw *hw, enum ice_block blk,
+                       struct LIST_HEAD_TYPE *chs, u16 *vsig);
+
+enum ice_status
+ice_vsig_add_mv_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig);
+enum ice_status ice_vsig_free(struct ice_hw *hw, enum ice_block blk, u16 vsig);
+enum ice_status
+ice_vsig_add_mv_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig);
+enum ice_status
+ice_vsig_remove_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig);
+enum ice_status
+ice_add_prof(struct ice_hw *hw, enum ice_block blk, u64 id, u8 ptypes[],
+            struct ice_fv_word *es);
+struct ice_prof_map *
+ice_search_prof_id(struct ice_hw *hw, enum ice_block blk, u64 id);
+enum ice_status
+ice_add_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl);
+enum ice_status
+ice_rem_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl);
+struct ice_prof_map *
+ice_set_prof_context(struct ice_hw *hw, enum ice_block blk, u64 id, u64 cntxt);
+struct ice_prof_map *
+ice_get_prof_context(struct ice_hw *hw, enum ice_block blk, u64 id, u64 *cntxt);
+enum ice_status
+ice_copy_and_init_pkg(struct ice_hw *hw, const u8 *buf, u32 len);
+enum ice_status ice_init_hw_tbls(struct ice_hw *hw);
+void ice_free_seg(struct ice_hw *hw);
+void ice_free_hw_tbls(struct ice_hw *hw);
+enum ice_status
+ice_add_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi[], u8 count,
+            u64 id);
+enum ice_status
+ice_rem_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi[], u8 count,
+            u64 id);
+enum ice_status
+ice_rem_prof(struct ice_hw *hw, enum ice_block blk, u64 id);
+
+enum ice_status
+ice_set_key(u8 *key, u16 size, u8 *val, u8 *upd, u8 *dc, u8 *nm, u16 off,
+           u16 len);
+#endif /* _ICE_FLEX_PIPE_H_ */
index 84a38cb..9f98f4f 100644 (file)
@@ -16,4 +16,690 @@ struct ice_fv {
        struct ice_fv_word ew[ICE_MAX_FV_WORDS];
 };
 
+/* Package and segment headers and tables */
+struct ice_pkg_hdr {
+       struct ice_pkg_ver format_ver;
+       __le32 seg_count;
+       __le32 seg_offset[1];
+};
+
+/* generic segment */
+struct ice_generic_seg_hdr {
+#define SEGMENT_TYPE_METADATA  0x00000001
+#define SEGMENT_TYPE_ICE       0x00000010
+       __le32 seg_type;
+       struct ice_pkg_ver seg_ver;
+       __le32 seg_size;
+       char seg_name[ICE_PKG_NAME_SIZE];
+};
+
+/* ice specific segment */
+
+union ice_device_id {
+       struct {
+               __le16 device_id;
+               __le16 vendor_id;
+       } dev_vend_id;
+       __le32 id;
+};
+
+struct ice_device_id_entry {
+       union ice_device_id device;
+       union ice_device_id sub_device;
+};
+
+struct ice_seg {
+       struct ice_generic_seg_hdr hdr;
+       __le32 device_table_count;
+       struct ice_device_id_entry device_table[1];
+};
+
+struct ice_nvm_table {
+       __le32 table_count;
+       __le32 vers[1];
+};
+
+struct ice_buf {
+#define ICE_PKG_BUF_SIZE       4096
+       u8 buf[ICE_PKG_BUF_SIZE];
+};
+
+struct ice_buf_table {
+       __le32 buf_count;
+       struct ice_buf buf_array[1];
+};
+
+/* global metadata specific segment */
+struct ice_global_metadata_seg {
+       struct ice_generic_seg_hdr hdr;
+       struct ice_pkg_ver pkg_ver;
+       __le32 track_id;
+       char pkg_name[ICE_PKG_NAME_SIZE];
+};
+
+#define ICE_MIN_S_OFF          12
+#define ICE_MAX_S_OFF          4095
+#define ICE_MIN_S_SZ           1
+#define ICE_MAX_S_SZ           4084
+
+/* section information */
+struct ice_section_entry {
+       __le32 type;
+       __le16 offset;
+       __le16 size;
+};
+
+#define ICE_MIN_S_COUNT                1
+#define ICE_MAX_S_COUNT                511
+#define ICE_MIN_S_DATA_END     12
+#define ICE_MAX_S_DATA_END     4096
+
+#define ICE_METADATA_BUF       0x80000000
+
+struct ice_buf_hdr {
+       __le16 section_count;
+       __le16 data_end;
+       struct ice_section_entry section_entry[1];
+};
+
+#define ICE_MAX_ENTRIES_IN_BUF(hd_sz, ent_sz) ((ICE_PKG_BUF_SIZE - \
+       sizeof(struct ice_buf_hdr) - (hd_sz)) / (ent_sz))
+
+/* ice package section IDs */
+#define ICE_SID_XLT0_SW                        10
+#define ICE_SID_XLT_KEY_BUILDER_SW     11
+#define ICE_SID_XLT1_SW                        12
+#define ICE_SID_XLT2_SW                        13
+#define ICE_SID_PROFID_TCAM_SW         14
+#define ICE_SID_PROFID_REDIR_SW                15
+#define ICE_SID_FLD_VEC_SW             16
+#define ICE_SID_CDID_KEY_BUILDER_SW    17
+#define ICE_SID_CDID_REDIR_SW          18
+
+#define ICE_SID_XLT0_ACL               20
+#define ICE_SID_XLT_KEY_BUILDER_ACL    21
+#define ICE_SID_XLT1_ACL               22
+#define ICE_SID_XLT2_ACL               23
+#define ICE_SID_PROFID_TCAM_ACL                24
+#define ICE_SID_PROFID_REDIR_ACL       25
+#define ICE_SID_FLD_VEC_ACL            26
+#define ICE_SID_CDID_KEY_BUILDER_ACL   27
+#define ICE_SID_CDID_REDIR_ACL         28
+
+#define ICE_SID_XLT0_FD                        30
+#define ICE_SID_XLT_KEY_BUILDER_FD     31
+#define ICE_SID_XLT1_FD                        32
+#define ICE_SID_XLT2_FD                        33
+#define ICE_SID_PROFID_TCAM_FD         34
+#define ICE_SID_PROFID_REDIR_FD                35
+#define ICE_SID_FLD_VEC_FD             36
+#define ICE_SID_CDID_KEY_BUILDER_FD    37
+#define ICE_SID_CDID_REDIR_FD          38
+
+#define ICE_SID_XLT0_RSS               40
+#define ICE_SID_XLT_KEY_BUILDER_RSS    41
+#define ICE_SID_XLT1_RSS               42
+#define ICE_SID_XLT2_RSS               43
+#define ICE_SID_PROFID_TCAM_RSS                44
+#define ICE_SID_PROFID_REDIR_RSS       45
+#define ICE_SID_FLD_VEC_RSS            46
+#define ICE_SID_CDID_KEY_BUILDER_RSS   47
+#define ICE_SID_CDID_REDIR_RSS         48
+
+#define ICE_SID_RXPARSER_CAM           50
+#define ICE_SID_RXPARSER_NOMATCH_CAM   51
+#define ICE_SID_RXPARSER_IMEM          52
+#define ICE_SID_RXPARSER_XLT0_BUILDER  53
+#define ICE_SID_RXPARSER_NODE_PTYPE    54
+#define ICE_SID_RXPARSER_MARKER_PTYPE  55
+#define ICE_SID_RXPARSER_BOOST_TCAM    56
+#define ICE_SID_RXPARSER_PROTO_GRP     57
+#define ICE_SID_RXPARSER_METADATA_INIT 58
+#define ICE_SID_RXPARSER_XLT0          59
+
+#define ICE_SID_TXPARSER_CAM           60
+#define ICE_SID_TXPARSER_NOMATCH_CAM   61
+#define ICE_SID_TXPARSER_IMEM          62
+#define ICE_SID_TXPARSER_XLT0_BUILDER  63
+#define ICE_SID_TXPARSER_NODE_PTYPE    64
+#define ICE_SID_TXPARSER_MARKER_PTYPE  65
+#define ICE_SID_TXPARSER_BOOST_TCAM    66
+#define ICE_SID_TXPARSER_PROTO_GRP     67
+#define ICE_SID_TXPARSER_METADATA_INIT 68
+#define ICE_SID_TXPARSER_XLT0          69
+
+#define ICE_SID_RXPARSER_INIT_REDIR    70
+#define ICE_SID_TXPARSER_INIT_REDIR    71
+#define ICE_SID_RXPARSER_MARKER_GRP    72
+#define ICE_SID_TXPARSER_MARKER_GRP    73
+#define ICE_SID_RXPARSER_LAST_PROTO    74
+#define ICE_SID_TXPARSER_LAST_PROTO    75
+#define ICE_SID_RXPARSER_PG_SPILL      76
+#define ICE_SID_TXPARSER_PG_SPILL      77
+#define ICE_SID_RXPARSER_NOMATCH_SPILL 78
+#define ICE_SID_TXPARSER_NOMATCH_SPILL 79
+
+#define ICE_SID_XLT0_PE                        80
+#define ICE_SID_XLT_KEY_BUILDER_PE     81
+#define ICE_SID_XLT1_PE                        82
+#define ICE_SID_XLT2_PE                        83
+#define ICE_SID_PROFID_TCAM_PE         84
+#define ICE_SID_PROFID_REDIR_PE                85
+#define ICE_SID_FLD_VEC_PE             86
+#define ICE_SID_CDID_KEY_BUILDER_PE    87
+#define ICE_SID_CDID_REDIR_PE          88
+
+/* Label Metadata section IDs */
+#define ICE_SID_LBL_FIRST              0x80000010
+#define ICE_SID_LBL_RXPARSER_IMEM      0x80000010
+#define ICE_SID_LBL_TXPARSER_IMEM      0x80000011
+#define ICE_SID_LBL_RESERVED_12                0x80000012
+#define ICE_SID_LBL_RESERVED_13                0x80000013
+#define ICE_SID_LBL_RXPARSER_MARKER    0x80000014
+#define ICE_SID_LBL_TXPARSER_MARKER    0x80000015
+#define ICE_SID_LBL_PTYPE              0x80000016
+#define ICE_SID_LBL_PROTOCOL_ID                0x80000017
+#define ICE_SID_LBL_RXPARSER_TMEM      0x80000018
+#define ICE_SID_LBL_TXPARSER_TMEM      0x80000019
+#define ICE_SID_LBL_RXPARSER_PG                0x8000001A
+#define ICE_SID_LBL_TXPARSER_PG                0x8000001B
+#define ICE_SID_LBL_RXPARSER_M_TCAM    0x8000001C
+#define ICE_SID_LBL_TXPARSER_M_TCAM    0x8000001D
+#define ICE_SID_LBL_SW_PROFID_TCAM     0x8000001E
+#define ICE_SID_LBL_ACL_PROFID_TCAM    0x8000001F
+#define ICE_SID_LBL_PE_PROFID_TCAM     0x80000020
+#define ICE_SID_LBL_RSS_PROFID_TCAM    0x80000021
+#define ICE_SID_LBL_FD_PROFID_TCAM     0x80000022
+#define ICE_SID_LBL_FLAG               0x80000023
+#define ICE_SID_LBL_REG                        0x80000024
+#define ICE_SID_LBL_SW_PTG             0x80000025
+#define ICE_SID_LBL_ACL_PTG            0x80000026
+#define ICE_SID_LBL_PE_PTG             0x80000027
+#define ICE_SID_LBL_RSS_PTG            0x80000028
+#define ICE_SID_LBL_FD_PTG             0x80000029
+#define ICE_SID_LBL_SW_VSIG            0x8000002A
+#define ICE_SID_LBL_ACL_VSIG           0x8000002B
+#define ICE_SID_LBL_PE_VSIG            0x8000002C
+#define ICE_SID_LBL_RSS_VSIG           0x8000002D
+#define ICE_SID_LBL_FD_VSIG            0x8000002E
+#define ICE_SID_LBL_PTYPE_META         0x8000002F
+#define ICE_SID_LBL_SW_PROFID          0x80000030
+#define ICE_SID_LBL_ACL_PROFID         0x80000031
+#define ICE_SID_LBL_PE_PROFID          0x80000032
+#define ICE_SID_LBL_RSS_PROFID         0x80000033
+#define ICE_SID_LBL_FD_PROFID          0x80000034
+#define ICE_SID_LBL_RXPARSER_MARKER_GRP        0x80000035
+#define ICE_SID_LBL_TXPARSER_MARKER_GRP        0x80000036
+#define ICE_SID_LBL_RXPARSER_PROTO     0x80000037
+#define ICE_SID_LBL_TXPARSER_PROTO     0x80000038
+/* The following define MUST be updated to reflect the last label section ID */
+#define ICE_SID_LBL_LAST               0x80000038
+
+enum ice_block {
+       ICE_BLK_SW = 0,
+       ICE_BLK_ACL,
+       ICE_BLK_FD,
+       ICE_BLK_RSS,
+       ICE_BLK_PE,
+       ICE_BLK_COUNT
+};
+
+enum ice_sect {
+       ICE_XLT0 = 0,
+       ICE_XLT_KB,
+       ICE_XLT1,
+       ICE_XLT2,
+       ICE_PROF_TCAM,
+       ICE_PROF_REDIR,
+       ICE_VEC_TBL,
+       ICE_CDID_KB,
+       ICE_CDID_REDIR,
+       ICE_SECT_COUNT
+};
+
+/* Packet Type (PTYPE) values */
+#define ICE_PTYPE_MAC_PAY              1
+#define ICE_PTYPE_IPV4FRAG_PAY         22
+#define ICE_PTYPE_IPV4_PAY             23
+#define ICE_PTYPE_IPV4_UDP_PAY         24
+#define ICE_PTYPE_IPV4_TCP_PAY         26
+#define ICE_PTYPE_IPV4_SCTP_PAY                27
+#define ICE_PTYPE_IPV4_ICMP_PAY                28
+#define ICE_PTYPE_IPV6FRAG_PAY         88
+#define ICE_PTYPE_IPV6_PAY             89
+#define ICE_PTYPE_IPV6_UDP_PAY         90
+#define ICE_PTYPE_IPV6_TCP_PAY         92
+#define ICE_PTYPE_IPV6_SCTP_PAY                93
+#define ICE_PTYPE_IPV6_ICMP_PAY                94
+
+/* Packet Type Groups (PTG) - Inner Most fields (IM) */
+#define ICE_PTG_IM_IPV4_TCP            16
+#define ICE_PTG_IM_IPV4_UDP            17
+#define ICE_PTG_IM_IPV4_SCTP           18
+#define ICE_PTG_IM_IPV4_PAY            20
+#define ICE_PTG_IM_IPV4_OTHER          21
+#define ICE_PTG_IM_IPV6_TCP            32
+#define ICE_PTG_IM_IPV6_UDP            33
+#define ICE_PTG_IM_IPV6_SCTP           34
+#define ICE_PTG_IM_IPV6_OTHER          37
+#define ICE_PTG_IM_L2_OTHER            67
+
+struct ice_flex_fields {
+       union {
+               struct {
+                       u8 src_ip;
+                       u8 dst_ip;
+                       u8 flow_label;  /* valid for IPv6 only */
+               } ip_fields;
+
+               struct {
+                       u8 src_prt;
+                       u8 dst_prt;
+               } tcp_udp_fields;
+
+               struct {
+                       u8 src_ip;
+                       u8 dst_ip;
+                       u8 src_prt;
+                       u8 dst_prt;
+               } ip_tcp_udp_fields;
+
+               struct {
+                       u8 src_prt;
+                       u8 dst_prt;
+                       u8 flow_label;  /* valid for IPv6 only */
+                       u8 spi;
+               } ip_esp_fields;
+
+               struct {
+                       u32 offset;
+                       u32 length;
+               } off_len;
+       } fields;
+};
+
+#define ICE_XLT1_DFLT_GRP      0
+#define ICE_XLT1_TABLE_SIZE    1024
+
+/* package labels */
+struct ice_label {
+       __le16 value;
+#define ICE_PKG_LABEL_SIZE     64
+       char name[ICE_PKG_LABEL_SIZE];
+};
+
+struct ice_label_section {
+       __le16 count;
+       struct ice_label label[1];
+};
+
+#define ICE_MAX_LABELS_IN_BUF ICE_MAX_ENTRIES_IN_BUF( \
+       sizeof(struct ice_label_section) - sizeof(struct ice_label), \
+       sizeof(struct ice_label))
+
+struct ice_sw_fv_section {
+       __le16 count;
+       __le16 base_offset;
+       struct ice_fv fv[1];
+};
+
+struct ice_sw_fv_list_entry {
+       struct LIST_ENTRY_TYPE list_entry;
+       u32 profile_id;
+       struct ice_fv *fv_ptr;
+};
+
+#pragma pack(1)
+/* The BOOST tcam stores the match packet header in reverse order, meaning
+ * the fields are reversed; in addition, this means that the normally big endian
+ * fields of the packet are now little endian.
+ */
+struct ice_boost_key_value {
+#define ICE_BOOST_REMAINING_HV_KEY     15
+       u8 remaining_hv_key[ICE_BOOST_REMAINING_HV_KEY];
+       __le16 hv_dst_port_key;
+       __le16 hv_src_port_key;
+       u8 tcam_search_key;
+};
+
+#pragma pack()
+
+struct ice_boost_key {
+       struct ice_boost_key_value key;
+       struct ice_boost_key_value key2;
+};
+
+/* package Boost TCAM entry */
+struct ice_boost_tcam_entry {
+       __le16 addr;
+       __le16 reserved;
+       /* break up the 40 bytes of key into different fields */
+       struct ice_boost_key key;
+       u8 boost_hit_index_group;
+       /* The following contains bitfields which are not on byte boundaries.
+        * These fields are currently unused by driver software.
+        */
+#define ICE_BOOST_BIT_FIELDS           43
+       u8 bit_fields[ICE_BOOST_BIT_FIELDS];
+};
+
+struct ice_boost_tcam_section {
+       __le16 count;
+       __le16 reserved;
+       struct ice_boost_tcam_entry tcam[1];
+};
+
+#define ICE_MAX_BST_TCAMS_IN_BUF ICE_MAX_ENTRIES_IN_BUF( \
+       sizeof(struct ice_boost_tcam_section) - \
+       sizeof(struct ice_boost_tcam_entry), \
+       sizeof(struct ice_boost_tcam_entry))
+
+#pragma pack(1)
+struct ice_xlt1_section {
+       __le16 count;
+       __le16 offset;
+       u8 value[1];
+};
+
+#pragma pack()
+
+#define ICE_XLT1_SIZE(n)       (sizeof(struct ice_xlt1_section) + \
+                                (sizeof(u8) * ((n) - 1)))
+
+struct ice_xlt2_section {
+       __le16 count;
+       __le16 offset;
+       __le16 value[1];
+};
+
+#define ICE_XLT2_SIZE(n)       (sizeof(struct ice_xlt2_section) + \
+                                (sizeof(u16) * ((n) - 1)))
+
+struct ice_prof_redir_section {
+       __le16 count;
+       __le16 offset;
+       u8 redir_value[1];
+};
+
+#define ICE_PROF_REDIR_SIZE(n) (sizeof(struct ice_prof_redir_section) + \
+                                (sizeof(u8) * ((n) - 1)))
+
+/* package buffer building */
+
+struct ice_buf_build {
+       struct ice_buf buf;
+       u16 reserved_section_table_entries;
+};
+
+struct ice_pkg_enum {
+       struct ice_buf_table *buf_table;
+       u32 buf_idx;
+
+       u32 type;
+       struct ice_buf_hdr *buf;
+       u32 sect_idx;
+       void *sect;
+       u32 sect_type;
+
+       u32 entry_idx;
+       void *(*handler)(u32 sect_type, void *section, u32 index, u32 *offset);
+};
+
+/* Tunnel enabling */
+
+enum ice_tunnel_type {
+       TNL_VXLAN = 0,
+       TNL_GTPC,
+       TNL_GTPC_TEID,
+       TNL_GTPU,
+       TNL_GTPU_TEID,
+       TNL_VXLAN_GPE,
+       TNL_GENEVE,
+       TNL_NAT,
+       TNL_ROCE_V2,
+       TNL_MPLSO_UDP,
+       TNL_UDP2_END,
+       TNL_UPD_END,
+       TNL_LAST = 0xFF,
+       TNL_ALL = 0xFF,
+};
+
+struct ice_tunnel_type_scan {
+       enum ice_tunnel_type type;
+       const char *label_prefix;
+};
+
+struct ice_tunnel_entry {
+       enum ice_tunnel_type type;
+       u8 valid;
+       u8 in_use;
+       u8 marked;
+       u16 boost_addr;
+       u16 port;
+       struct ice_boost_tcam_entry *boost_entry;
+};
+
+#define ICE_TUNNEL_MAX_ENTRIES 16
+
+struct ice_tunnel_table {
+       u16 count;
+       struct ice_tunnel_entry tbl[ICE_TUNNEL_MAX_ENTRIES];
+};
+
+struct ice_pkg_es {
+       __le16 count;
+       __le16 offset;
+       struct ice_fv_word es[1];
+};
+
+struct ice_es {
+       u32 sid;
+       u16 count;
+       u16 fvw;
+       u16 *ref_count;
+       u8 reverse; /* set to true to reverse FV order */
+       struct LIST_HEAD_TYPE prof_map;
+       struct ice_fv_word *t;
+       u8 *resource_used_hack; /* hack for testing */
+};
+
+/* PTYPE Group management */
+
+/* Note: XLT1 table takes 13-bit as input, and results in an 8-bit packet type
+ * group (PTG) ID as output.
+ *
+ * Note: PTG 0 is the default packet type group and it is assumed that all PTYPE
+ * are a part of this group until moved to a new PTG.
+ */
+#define ICE_DEFAULT_PTG        0
+
+struct ice_ptg_entry {
+       u8 in_use;
+       struct ice_ptg_ptype *first_ptype;
+};
+
+struct ice_ptg_ptype {
+       u8 ptg;
+       struct ice_ptg_ptype *next_ptype;
+};
+
+#define ICE_MAX_TCAM_PER_PROFILE       8
+#define ICE_MAX_PTYPE_PER_PROFILE      8
+
+struct ice_prof_map {
+       struct LIST_ENTRY_TYPE list;
+       u64 profile_cookie;
+       u64 context;
+       u8 prof_id;
+       u8 ptype_count;
+       u8 ptype[ICE_MAX_PTYPE_PER_PROFILE];
+};
+
+#define ICE_INVALID_TCAM       0xFFFF
+
+struct ice_tcam_inf {
+       u8 ptg;
+       u8 prof_id;
+       u16 tcam_idx;
+       u8 in_use;
+};
+
+struct ice_vsig_prof {
+       struct LIST_ENTRY_TYPE list;
+       u64 profile_cookie;
+       u8 prof_id;
+       u8 tcam_count;
+       struct ice_tcam_inf tcam[ICE_MAX_TCAM_PER_PROFILE];
+};
+
+struct ice_vsig_entry {
+       u8 in_use;
+       struct LIST_HEAD_TYPE prop_lst;
+       struct ice_vsig_vsi *first_vsi;
+};
+
+struct ice_vsig_vsi {
+       u16 changed;
+       u16 vsig;
+       u32 prop_mask;
+       struct ice_vsig_vsi *next_vsi;
+};
+
+#define ICE_XLT1_CNT   1024
+#define ICE_MAX_PTGS   256
+
+/* XLT1 Table */
+struct ice_xlt1 {
+       u32 sid;
+       u16 count;
+       struct ice_ptg_entry *ptg_tbl;
+       struct ice_ptg_ptype *ptypes;
+       u8 *t;
+};
+
+#define ICE_XLT2_CNT   768
+#define ICE_MAX_VSIGS  768
+
+/* Vsig bit layout:
+ * [0:12]: incremental vsig index 1 to ICE_MAX_VSIGS
+ * [13:15]: pf number of device
+ */
+#define ICE_VSIG_IDX_M (0x1FFF)
+#define ICE_PF_NUM_S   13
+#define ICE_PF_NUM_M   (0x07 << ICE_PF_NUM_S)
+#define ICE_VSIG_VALUE(vsig, pf_id) \
+       (u16)((((u16)(vsig)) & ICE_VSIG_IDX_M) | \
+             (((u16)(pf_id) << ICE_PF_NUM_S) & ICE_PF_NUM_M))
+#define ICE_DEFAULT_VSIG       0
+
+/* XLT2 Table */
+struct ice_xlt2 {
+       u32 sid;
+       u16 count;
+       struct ice_vsig_entry *vsig_tbl;
+       struct ice_vsig_vsi *vsis;
+       u16 *t;
+};
+
+/* Extraction sequence - list of match fields:
+ * protocol id, offset, profile length
+ */
+union ice_match_fld {
+       struct {
+               u8 prot_id;
+               u8 offset;
+               u8 length;
+               u8 reserved; /* must be zero */
+       } fld;
+       u32 val;
+};
+
+#define ICE_MATCH_LIST_SZ      20
+#pragma pack(1)
+struct ice_match {
+       u8 count;
+       union ice_match_fld list[ICE_MATCH_LIST_SZ];
+};
+
+/* Profile ID Management */
+struct ice_prof_id_key {
+       __le16 flags;
+       u8 xlt1;
+       __le16 xlt2_cdid;
+};
+
+/* Keys are made up of two values, each one-half the size of the key.
+ * For TCAM, the entire key is 80 bits wide (or 2, 40-bit wide values)
+ */
+#define ICE_TCAM_KEY_VAL_SZ    5
+#define ICE_TCAM_KEY_SZ                (2 * ICE_TCAM_KEY_VAL_SZ)
+
+struct ice_prof_tcam_entry {
+       __le16 addr;
+       u8 key[ICE_TCAM_KEY_SZ];
+       u8 prof_id;
+};
+
+struct ice_prof_id_section {
+       __le16 count;
+       struct ice_prof_tcam_entry entry[1];
+};
+
+#pragma pack()
+
+struct ice_prof_tcam {
+       u32 sid;
+       u16 count;
+       u16 max_prof_id;
+       u8 cdid_bits; /* # cdid bits to use in key, 0, 2, 4, or 8 */
+       struct ice_prof_tcam_entry *t;
+       u8 *resource_used_hack;
+};
+
+struct ice_prof_redir {
+       u32 sid;
+       u16 count;
+       u8 *t;
+};
+
+/* Tables per block */
+struct ice_blk_info {
+       struct ice_xlt1 xlt1;
+       struct ice_xlt2 xlt2;
+       struct ice_prof_tcam prof;
+       struct ice_prof_redir prof_redir;
+       struct ice_es es;
+       u8 overwrite; /* set to true to allow overwrite of table entries */
+};
+
+enum ice_chg_type {
+       ICE_TCAM_NONE = 0,
+       ICE_PTG_ES_ADD,
+       ICE_TCAM_ADD,
+       ICE_TCAM_REM,
+       ICE_VSIG_ADD,
+       ICE_VSIG_REM,
+       ICE_VSI_MOVE,
+};
+
+struct ice_chs_chg {
+       struct LIST_ENTRY_TYPE list_entry;
+       enum ice_chg_type type;
+
+       u8 add_ptg;
+       u8 add_vsig;
+       u8 add_tcam_idx;
+       u8 add_prof;
+       u16 ptype;
+       u8 ptg;
+       u8 prof_id;
+       u16 vsi;
+       u16 vsig;
+       u16 orig_vsig;
+       u16 tcam_idx;
+       struct ice_prof_tcam_entry orig_ent;
+};
+
+#define ICE_FLOW_PTYPE_MAX             ICE_XLT1_CNT
+
 #endif /* _ICE_FLEX_TYPE_H_ */
index d7c12d1..17d79ba 100644 (file)
@@ -704,6 +704,42 @@ struct ice_hw {
 
        u8 ucast_shared;        /* true if VSIs can share unicast addr */
 
+       /* Active package version (currently active) */
+       struct ice_pkg_ver active_pkg_ver;
+       u8 active_pkg_name[ICE_PKG_NAME_SIZE];
+
+       /* Driver's package ver - (from the Metadata seg) */
+       struct ice_pkg_ver pkg_ver;
+       u8 pkg_name[ICE_PKG_NAME_SIZE];
+
+       /* Driver's Ice package version (from the Ice seg) */
+       struct ice_pkg_ver ice_pkg_ver;
+       u8 ice_pkg_name[ICE_PKG_NAME_SIZE];
+
+       /* Pointer to the ice segment */
+       struct ice_seg *seg;
+
+       /* Pointer to allocated copy of pkg memory */
+       u8 *pkg_copy;
+
+       /* tunneling info */
+       struct ice_tunnel_table tnl;
+
+       /* PTYPE group and XLT1 management */
+#define ICE_MAX_PTGS   256
+       struct ice_ptg_entry ptg_tbl[ICE_BLK_COUNT][ICE_MAX_PTGS];
+
+#define ICE_XLT1_CNT   1024
+       struct ice_ptg_ptype xlt1_tbl[ICE_BLK_COUNT][ICE_XLT1_CNT];
+#define ICE_PKG_FILENAME       "package_file"
+#define ICE_PKG_FILENAME_EXT   "pkg"
+#define ICE_PKG_FILE_MAJ_VER   1
+#define ICE_PKG_FILE_MIN_VER   0
+
+       /* HW block tables */
+       struct ice_blk_info blk[ICE_BLK_COUNT];
+       struct ice_lock fl_profs_locks[ICE_BLK_COUNT];  /* lock fltr profiles */
+       struct LIST_HEAD_TYPE fl_profs[ICE_BLK_COUNT];
 };
 
 /* Statistics collected by each port, VSI, VEB, and S-channel */
index 0cfc8cd..77a9802 100644 (file)
@@ -7,6 +7,7 @@ sources = [
        'ice_sched.c',
        'ice_switch.c',
        'ice_nvm.c',
+       'ice_flex_pipe.c',
 ]
 
 error_cflags = ['-Wno-unused-value',