scsi: lpfc: Revise lpfc_prep_embed_io routine with proper endian macro usages

On big endian architectures, it is possible to run into a memory out of
bounds pointer dereference when FCP targets are zoned.

In lpfc_prep_embed_io, the memcpy(ptr, fcp_cmnd, sgl->sge_len) is
referencing a little endian formatted sgl->sge_len value.  So, the memcpy
can cause big endian systems to crash.

Redefine the *sgl ptr as a struct sli4_sge_le to make it clear that we are
referring to a little endian formatted data structure.  And, update the
routine with proper le32_to_cpu macro usages.

Fixes: af20bb73 ("scsi: lpfc: Add support for 32 byte CDBs")
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://lore.kernel.org/r/20240628172011.25921-8-justintee8345@gmail.comSigned-off-by: Martin K. Petersen <martin.petersen@oracle.com>

drivers/scsi/lpfc/lpfc_sli.c

@@ -10579,10 +10579,11 @@ lpfc_prep_embed_io(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_cmd)
 {
 	struct lpfc_iocbq *piocb = &lpfc_cmd->cur_iocbq;
 	union lpfc_wqe128 *wqe = &lpfc_cmd->cur_iocbq.wqe;
-	struct sli4_sge *sgl;
+	struct sli4_sge_le *sgl;
+	u32 type_size;
 
 	/* 128 byte wqe support here */
-	sgl = (struct sli4_sge *)lpfc_cmd->dma_sgl;
+	sgl = (struct sli4_sge_le *)lpfc_cmd->dma_sgl;
 
 	if (phba->fcp_embed_io) {
 		struct fcp_cmnd *fcp_cmnd;
@@ -10591,9 +10592,9 @@ lpfc_prep_embed_io(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_cmd)
 		fcp_cmnd = lpfc_cmd->fcp_cmnd;
 
 		/* Word 0-2 - FCP_CMND */
-		wqe->generic.bde.tus.f.bdeFlags =
-			BUFF_TYPE_BDE_IMMED;
-		wqe->generic.bde.tus.f.bdeSize = sgl->sge_len;
+		type_size = le32_to_cpu(sgl->sge_len);
+		type_size |= ULP_BDE64_TYPE_BDE_IMMED;
+		wqe->generic.bde.tus.w = type_size;
 		wqe->generic.bde.addrHigh = 0;
 		wqe->generic.bde.addrLow =  72;  /* Word 18 */
 
@@ -10602,13 +10603,13 @@ lpfc_prep_embed_io(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_cmd)
 
 		/* Word 18-29  FCP CMND Payload */
 		ptr = &wqe->words[18];
-		memcpy(ptr, fcp_cmnd, sgl->sge_len);
+		lpfc_sli_pcimem_bcopy(fcp_cmnd, ptr, le32_to_cpu(sgl->sge_len));
 	} else {
 		/* Word 0-2 - Inline BDE */
 		wqe->generic.bde.tus.f.bdeFlags =  BUFF_TYPE_BDE_64;
-		wqe->generic.bde.tus.f.bdeSize = sgl->sge_len;
-		wqe->generic.bde.addrHigh = sgl->addr_hi;
-		wqe->generic.bde.addrLow =  sgl->addr_lo;
+		wqe->generic.bde.tus.f.bdeSize = le32_to_cpu(sgl->sge_len);
+		wqe->generic.bde.addrHigh = le32_to_cpu(sgl->addr_hi);
+		wqe->generic.bde.addrLow = le32_to_cpu(sgl->addr_lo);
 
 		/* Word 10 */
 		bf_set(wqe_dbde, &wqe->generic.wqe_com, 1);