[Skiboot] [PATCH V5 2/2] core/pldm: Register PLDM as a blocklevel device

Christophe Lombard clombard at linux.ibm.com
Tue Aug 29 19:23:54 AEST 2023


In the same way that ipmi-hiomap implements the PNOR access control
protocol, this patch allows to "virtualize" the content of a BMC flash
based on lid files.
Previously, flash PNOR partitions were viewed this way:
  partitionXX=NAME, start address, end address, flags

The content of each partition is now stored in a lid file. In order to
continue to use the libflash library, we manually fill in the contents of
a fake flash header when accessing offset 0. This reproduces the behavior
via ipmi-hiomap of reading the flash header on the BMC.

For the reading and writing of BMC lids files, we convert the virtual
addresses of these 'fake' partitions by identifying: lid id.

Signed-off-by: Christophe Lombard <clombard at linux.ibm.com>
---
 core/pldm/pldm-lid-files.c | 247 +++++++++++++++++++++++++++++++++++++
 include/pldm.h             |   5 +
 2 files changed, 252 insertions(+)

diff --git a/core/pldm/pldm-lid-files.c b/core/pldm/pldm-lid-files.c
index a46a73164..19f0a9e8e 100644
--- a/core/pldm/pldm-lid-files.c
+++ b/core/pldm/pldm-lid-files.c
@@ -25,6 +25,14 @@ struct pldm_lid {
 
 static LIST_HEAD(lid_files);
 
+struct pldm_ctx_data {
+	/* Members protected by the blocklevel lock */
+	struct blocklevel_device bl;
+	uint32_t total_size;
+	uint32_t erase_granule;
+	struct lock lock;
+};
+
 #define MEGABYTE (1024*1024)
 
 /*
@@ -34,6 +42,12 @@ static LIST_HEAD(lid_files);
  */
 #define VMM_SIZE_RESERVED_PER_SECTION (32 * MEGABYTE)
 
+#define ERASE_GRANULE_DEF 0x1000
+
+/* 'fake' header flash */
+struct __ffs_hdr *raw_hdr;
+size_t raw_hdr_size;
+
 /*
  * Print the attributes of lid files.
  */
@@ -159,8 +173,221 @@ out:
 	return rc;
 }
 
+static uint32_t checksum(void *data, size_t size)
+{
+	uint32_t i, csum = 0;
+
+	for (i = csum = 0; i < (size/4); i++)
+		csum ^= ((uint32_t *)data)[i];
+	return csum;
+}
+
+/* Helper functions for typesafety and size safety */
+static uint32_t hdr_checksum(struct __ffs_hdr *hdr)
+{
+	return checksum(hdr, sizeof(struct __ffs_hdr));
+}
+
+static uint32_t entry_checksum(struct __ffs_entry *ent)
+{
+	return checksum(ent, sizeof(struct __ffs_entry));
+}
+
+/*
+ * Fill __ffs structures in order to return a 'fake' header flash
+ */
+static int lid_ids_to_header_flash(void *buf, uint64_t len)
+{
+	struct __ffs_entry *entry;
+	struct pldm_lid *lid = NULL;
+	uint32_t count, part_id, i;
+	uint32_t block_size;
+
+	/* reading the flash header has already been requested */
+	if (raw_hdr) {
+		(raw_hdr_size < len) ? memcpy(buf, raw_hdr, raw_hdr_size) :
+				       memcpy(buf, raw_hdr, len);
+		return OPAL_SUCCESS;
+	}
+
+	/* number of lid files */
+	count = get_lids_count();
+
+	/* last member of struct __ffs_hdr is a flexible array member */
+	raw_hdr_size = sizeof(struct __ffs_hdr) + (count * sizeof(struct __ffs_entry));
+	raw_hdr = zalloc(raw_hdr_size);
+	if (!raw_hdr)
+		return OPAL_NO_MEM;
+
+	/* complete header flash
+	 * Represents the on flash layout of FFS structures
+	 * Note: Beware that the size of the partition table is in units of block_size
+	 *
+	 * @magic:		Eye catcher/corruption detector
+	 * @version:		Version of the structure
+	 * @size:		Size of partition table (in block_size)
+	 * @entry_size:		Size of struct __ffs_entry element (in bytes)
+	 * @entry_count:	Number of struct __ffs_entry elements in @entries array
+	 * @block_size:		Size of block on device (in bytes)
+	 * @block_count:	Number of blocks on device
+	 * @checksum:		Header checksum
+	 */
+	/* size of the cached map: block_size * raw_hdr->size
+	 * raw_hdr->size = 0x3: we take a little margin if the number
+	 * of element would increase
+	 */
+	block_size = ERASE_GRANULE_DEF;
+
+	raw_hdr->magic = cpu_to_be32(FFS_MAGIC);
+	raw_hdr->version = cpu_to_be32(FFS_VERSION_1);
+	raw_hdr->size = cpu_to_be32(0x3);
+	raw_hdr->entry_size = cpu_to_be32(sizeof(struct __ffs_entry));
+	raw_hdr->entry_count = cpu_to_be32(count);
+	raw_hdr->block_size = cpu_to_be32(block_size);
+	raw_hdr->block_count = cpu_to_be32(0x4000); /* value from IPMI/PNOR protocol */
+	raw_hdr->checksum = hdr_checksum(raw_hdr);
+
+	lid = list_top(&lid_files, struct pldm_lid, list);
+	part_id = 1;
+
+	for (i = 0; i < count; i++) {
+		entry = &raw_hdr->entries[i];
+
+		memcpy(entry->name, lid->name, sizeof(entry->name));
+		entry->name[FFS_PART_NAME_MAX] = '\0';
+		entry->base = cpu_to_be32(lid->start / block_size);
+		entry->size = cpu_to_be32(lid->length / block_size);
+		entry->pid = cpu_to_be32(FFS_PID_TOPLEVEL);
+		entry->id = cpu_to_be32(part_id);
+		entry->type = cpu_to_be32(0x1);
+		entry->flags = cpu_to_be32(0x0);
+		entry->actual = cpu_to_be32(lid->length);
+		entry->checksum = entry_checksum(entry);
+
+		lid = list_next(&lid_files, lid, list);
+		part_id++;
+	}
+
+	/* fill in rquester buffer */
+	(raw_hdr_size < len) ? memcpy(buf, raw_hdr, raw_hdr_size) :
+			       memcpy(buf, raw_hdr, len);
+
+	return OPAL_SUCCESS;
+}
+
+/*
+ * Search lid member from the virtual address.
+ */
+static int vaddr_to_lid_id(uint64_t pos, uint32_t *start, uint32_t *handle,
+			   uint32_t *length)
+{
+	struct pldm_lid *lid = NULL;
+
+	list_for_each(&lid_files, lid, list) {
+		if ((pos >= lid->start) && (pos < lid->start + VMM_SIZE_RESERVED_PER_SECTION)) {
+			*start = lid->start;
+			*handle = lid->handle;
+			*length = lid->length;
+			return OPAL_SUCCESS;
+		}
+	}
+
+	return OPAL_PARAMETER;
+}
+
+static int lid_files_read(struct blocklevel_device *bl __unused,
+			  uint64_t pos, void *buf, uint64_t len)
+{
+	uint32_t lid_start, lid_handle, lid_length;
+	int rc = OPAL_SUCCESS;
+	uint64_t offset;
+
+	/* LPC is only 32bit */
+	if (pos > UINT_MAX || (pos + len) > UINT_MAX)
+		return FLASH_ERR_PARM_ERROR;
+
+	prlog(PR_TRACE, "lid files read at 0x%llx for 0x%llx\n",
+			pos, len);
+
+	if ((pos == 0) || (pos <= (ERASE_GRANULE_DEF * 0x3))) {
+		/* return a 'fake' header flash or cached map */
+		rc = lid_ids_to_header_flash(buf, len);
+	} else {
+		/* convert offset to lid id */
+		rc = vaddr_to_lid_id(pos, &lid_start,
+				     &lid_handle, &lid_length);
+		if (rc)
+			return rc;
+
+		/* read lid file */
+		offset = pos - lid_start;
+		rc = pldm_file_io_read_file(lid_handle, lid_length,
+					    offset, buf, len);
+	}
+
+	return rc;
+}
+
+static int lid_files_write(struct blocklevel_device *bl __unused,
+			   uint64_t pos, const void *buf __unused,
+			   uint64_t len)
+{
+	prlog(PR_ERR, "lid files writes at 0x%llx for 0x%llx\n",
+		       pos, len);
+	return OPAL_UNSUPPORTED;
+}
+
+static int lid_files_erase(struct blocklevel_device *bl __unused,
+			   uint64_t pos, uint64_t len)
+{
+
+	prlog(PR_ERR, "lid files erase at 0x%llx for 0x%llx\n",
+		       pos, len);
+	return OPAL_UNSUPPORTED;
+}
+
+static int get_lid_files_info(struct blocklevel_device *bl,
+			      const char **name, uint64_t *total_size,
+			      uint32_t *erase_granule)
+{
+	struct pldm_ctx_data *ctx;
+
+	ctx = container_of(bl, struct pldm_ctx_data, bl);
+	ctx->bl.erase_mask = ctx->erase_granule - 1;
+
+	if (name)
+		*name = NULL;
+	if (total_size)
+		*total_size = ctx->total_size;
+	if (erase_granule)
+		*erase_granule = ctx->erase_granule;
+
+	return OPAL_SUCCESS;
+}
+
+bool pldm_lid_files_exit(struct blocklevel_device *bl)
+{
+	struct pldm_ctx_data *ctx;
+	struct pldm_lid *lid, *tmp;
+
+	if (bl) {
+		ctx = container_of(bl, struct pldm_ctx_data, bl);
+		free(ctx);
+	}
+
+	/* free all lid entries */
+	list_for_each_safe(&lid_files, lid, tmp, list)
+		free(lid);
+
+	if (raw_hdr)
+		free(raw_hdr);
+
+	return true;
+}
+
 int pldm_lid_files_init(struct blocklevel_device **bl)
 {
+	struct pldm_ctx_data *ctx;
 	uint32_t lid_files_count;
 	int rc;
 
@@ -169,6 +396,18 @@ int pldm_lid_files_init(struct blocklevel_device **bl)
 
 	*bl = NULL;
 
+	ctx = zalloc(sizeof(struct pldm_ctx_data));
+	if (!ctx)
+		return FLASH_ERR_MALLOC_FAILED;
+
+	init_lock(&ctx->lock);
+
+	ctx->bl.read = &lid_files_read;
+	ctx->bl.write = &lid_files_write;
+	ctx->bl.erase = &lid_files_erase;
+	ctx->bl.get_info = &get_lid_files_info;
+	ctx->bl.exit = &pldm_lid_files_exit;
+
 	/* convert lid ids data to pnor structure */
 	rc = lid_ids_to_vaddr_mapping();
 	if (rc)
@@ -179,8 +418,16 @@ int pldm_lid_files_init(struct blocklevel_device **bl)
 	prlog(PR_NOTICE, "Number of lid files: %d\n", lid_files_count);
 	print_lid_files_attr();
 
+	ctx->total_size = lid_files_count * VMM_SIZE_RESERVED_PER_SECTION;
+	ctx->erase_granule = ERASE_GRANULE_DEF;
+
+	ctx->bl.keep_alive = 0;
+
+	*bl = &(ctx->bl);
+
 	return OPAL_SUCCESS;
 
 err:
+	free(ctx);
 	return rc;
 }
diff --git a/include/pldm.h b/include/pldm.h
index 9c5369edb..8622453be 100644
--- a/include/pldm.h
+++ b/include/pldm.h
@@ -43,4 +43,9 @@ int pldm_fru_dt_add_bmc_version(void);
  */
 int pldm_lid_files_init(struct blocklevel_device **bl);
 
+/**
+ * Remove lid ids data
+ */
+bool pldm_lid_files_exit(struct blocklevel_device *bl);
+
 #endif /* __PLDM_H__ */
-- 
2.41.0



More information about the Skiboot mailing list