diff mbox series

[v2,3/5] crypto: caam - avoid allocating memory at crypto request runtime for hash

Message ID 20230609103842.4159321-4-meenakshi.aggarwal@nxp.com (mailing list archive)
State Changes Requested
Delegated to: Herbert Xu
Headers show
Series Remove CRYPTO_ALG_ALLOCATES_MEMORY flag | expand

Commit Message

Meenakshi Aggarwal June 9, 2023, 10:38 a.m. UTC 
From: Meenakshi Aggarwal <meenakshi.aggarwal@nxp.com>

Remove CRYPTO_ALG_ALLOCATES_MEMORY flag and allocate the memory
needed by the driver, to fulfil a request, within the crypto
request object.
The extra size needed for base extended descriptor and link tables
is computed in frontend driver (caamhash) initialization and saved
in reqsize field that indicates how much memory could be needed per
request.

In reqsize we allocate memory for maximum 4 entries
for src and 4, aligned.
If the driver needs more than the 4 entries maximum, the memory
is dynamically allocated, at runtime.

Signed-off-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Signed-off-by: Meenakshi Aggarwal <meenakshi.aggarwal@nxp.com>
---
 drivers/crypto/caam/caamhash.c | 77 ++++++++++++++++++++++++----------
 1 file changed, 56 insertions(+), 21 deletions(-)
diff mbox series

Patch

diff --git a/drivers/crypto/caam/caamhash.c b/drivers/crypto/caam/caamhash.c
index 80deb003f0a5..75df8bca9ca9 100644
--- a/drivers/crypto/caam/caamhash.c
+++ b/drivers/crypto/caam/caamhash.c
@@ -535,6 +535,7 @@  static int acmac_setkey(struct crypto_ahash *ahash, const u8 *key,
  * @src_nents: number of segments in input scatterlist
  * @sec4_sg_bytes: length of dma mapped sec4_sg space
  * @bklog: stored to determine if the request needs backlog
+ * @free: stored to determine if ahash_edesc needs to be freed
  * @hw_desc: the h/w job descriptor followed by any referenced link tables
  * @sec4_sg: h/w link table
  */
@@ -543,6 +544,7 @@  struct ahash_edesc {
 	int src_nents;
 	int sec4_sg_bytes;
 	bool bklog;
+	bool free;
 	u32 hw_desc[DESC_JOB_IO_LEN_MAX / sizeof(u32)] ____cacheline_aligned;
 	struct sec4_sg_entry sec4_sg[];
 };
@@ -603,7 +605,8 @@  static inline void ahash_done_cpy(struct device *jrdev, u32 *desc, u32 err,
 
 	ahash_unmap_ctx(jrdev, edesc, req, digestsize, dir);
 	memcpy(req->result, state->caam_ctx, digestsize);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 
 	print_hex_dump_debug("ctx@"__stringify(__LINE__)": ",
 			     DUMP_PREFIX_ADDRESS, 16, 4, state->caam_ctx,
@@ -652,7 +655,8 @@  static inline void ahash_done_switch(struct device *jrdev, u32 *desc, u32 err,
 		ecode = caam_jr_strstatus(jrdev, err);
 
 	ahash_unmap_ctx(jrdev, edesc, req, ctx->ctx_len, dir);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 
 	scatterwalk_map_and_copy(state->buf, req->src,
 				 req->nbytes - state->next_buflen,
@@ -703,17 +707,28 @@  static struct ahash_edesc *ahash_edesc_alloc(struct ahash_request *req,
 					     dma_addr_t sh_desc_dma)
 {
 	struct crypto_ahash *ahash = crypto_ahash_reqtfm(req);
-	struct caam_hash_ctx *ctx = crypto_ahash_ctx_dma(ahash);
 	struct caam_hash_state *state = ahash_request_ctx_dma(req);
 	gfp_t flags = (req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP) ?
 		       GFP_KERNEL : GFP_ATOMIC;
 	struct ahash_edesc *edesc;
 	unsigned int sg_size = sg_num * sizeof(struct sec4_sg_entry);
-
-	edesc = kzalloc(sizeof(*edesc) + sg_size, flags);
-	if (!edesc) {
-		dev_err(ctx->jrdev, "could not allocate extended descriptor\n");
-		return NULL;
+	int edesc_size;
+
+	 /* Check if there's enough space for edesc saved in req */
+	edesc_size = sizeof(*edesc) + sg_size;
+	if (edesc_size > (crypto_ahash_reqsize(ahash) -
+			  sizeof(struct caam_hash_state))) {
+		/* allocate space for base edesc and link tables */
+		edesc = kzalloc(sizeof(*edesc) + sg_size, flags);
+		if (!edesc)
+			return ERR_PTR(-ENOMEM);
+		edesc->free = true;
+	} else {
+		/* get address for base edesc and link tables */
+		edesc = (struct ahash_edesc *)((u8 *)state +
+			 sizeof(struct caam_hash_state));
+		/* clear memory */
+		memset(edesc, 0, sizeof(*edesc));
 	}
 
 	state->edesc = edesc;
@@ -778,7 +793,8 @@  static int ahash_do_one_req(struct crypto_engine *engine, void *areq)
 
 	if (ret != -EINPROGRESS) {
 		ahash_unmap(jrdev, state->edesc, req, 0);
-		kfree(state->edesc);
+		if (state->edesc->free)
+			kfree(state->edesc);
 	} else {
 		ret = 0;
 	}
@@ -813,7 +829,8 @@  static int ahash_enqueue_req(struct device *jrdev,
 
 	if ((ret != -EINPROGRESS) && (ret != -EBUSY)) {
 		ahash_unmap_ctx(jrdev, edesc, req, dst_len, dir);
-		kfree(edesc);
+		if (edesc->free)
+			kfree(edesc);
 	}
 
 	return ret;
@@ -941,7 +958,8 @@  static int ahash_update_ctx(struct ahash_request *req)
 	return ret;
 unmap_ctx:
 	ahash_unmap_ctx(jrdev, edesc, req, ctx->ctx_len, DMA_BIDIRECTIONAL);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return ret;
 }
 
@@ -1002,7 +1020,8 @@  static int ahash_final_ctx(struct ahash_request *req)
 				 digestsize, DMA_BIDIRECTIONAL);
  unmap_ctx:
 	ahash_unmap_ctx(jrdev, edesc, req, digestsize, DMA_BIDIRECTIONAL);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return ret;
 }
 
@@ -1076,7 +1095,8 @@  static int ahash_finup_ctx(struct ahash_request *req)
 				 digestsize, DMA_BIDIRECTIONAL);
  unmap_ctx:
 	ahash_unmap_ctx(jrdev, edesc, req, digestsize, DMA_BIDIRECTIONAL);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return ret;
 }
 
@@ -1125,7 +1145,8 @@  static int ahash_digest(struct ahash_request *req)
 				  req->nbytes);
 	if (ret) {
 		ahash_unmap(jrdev, edesc, req, digestsize);
-		kfree(edesc);
+		if (edesc->free)
+			kfree(edesc);
 		return ret;
 	}
 
@@ -1134,7 +1155,8 @@  static int ahash_digest(struct ahash_request *req)
 	ret = map_seq_out_ptr_ctx(desc, jrdev, state, digestsize);
 	if (ret) {
 		ahash_unmap(jrdev, edesc, req, digestsize);
-		kfree(edesc);
+		if (edesc->free)
+			kfree(edesc);
 		return -ENOMEM;
 	}
 
@@ -1191,7 +1213,8 @@  static int ahash_final_no_ctx(struct ahash_request *req)
 				 digestsize, DMA_FROM_DEVICE);
  unmap:
 	ahash_unmap(jrdev, edesc, req, digestsize);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return -ENOMEM;
 }
 
@@ -1312,7 +1335,8 @@  static int ahash_update_no_ctx(struct ahash_request *req)
 	return ret;
  unmap_ctx:
 	ahash_unmap_ctx(jrdev, edesc, req, ctx->ctx_len, DMA_TO_DEVICE);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return ret;
 }
 
@@ -1387,7 +1411,8 @@  static int ahash_finup_no_ctx(struct ahash_request *req)
 				 digestsize, DMA_FROM_DEVICE);
  unmap:
 	ahash_unmap(jrdev, edesc, req, digestsize);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return -ENOMEM;
 
 }
@@ -1495,7 +1520,8 @@  static int ahash_update_first(struct ahash_request *req)
 	return ret;
  unmap_ctx:
 	ahash_unmap_ctx(jrdev, edesc, req, ctx->ctx_len, DMA_TO_DEVICE);
-	kfree(edesc);
+	if (edesc->free)
+		kfree(edesc);
 	return ret;
 }
 
@@ -1782,6 +1808,7 @@  static int caam_hash_cra_init(struct crypto_tfm *tfm)
 						      sh_desc_update);
 	dma_addr_t dma_addr;
 	struct caam_drv_private *priv;
+	int extra_reqsize = 0;
 
 	/*
 	 * Get a Job ring from Job Ring driver to ensure in-order
@@ -1862,7 +1889,15 @@  static int caam_hash_cra_init(struct crypto_tfm *tfm)
 
 	ctx->enginectx.op.do_one_request = ahash_do_one_req;
 
-	crypto_ahash_set_reqsize_dma(ahash, sizeof(struct caam_hash_state));
+	/* Compute extra space needed for base edesc and link tables */
+	extra_reqsize = sizeof(struct ahash_edesc) +
+			/* link tables for src:
+			 * 4 entries max + max 2 for remaining buf, aligned = 8
+			 */
+			(8 * sizeof(struct sec4_sg_entry));
+
+	crypto_ahash_set_reqsize_dma(ahash,
+				     sizeof(struct caam_hash_state) + extra_reqsize);
 
 	/*
 	 * For keyed hash algorithms shared descriptors
@@ -1937,7 +1972,7 @@  caam_hash_alloc(struct caam_hash_template *template,
 	alg->cra_priority = CAAM_CRA_PRIORITY;
 	alg->cra_blocksize = template->blocksize;
 	alg->cra_alignmask = 0;
-	alg->cra_flags = CRYPTO_ALG_ASYNC | CRYPTO_ALG_ALLOCATES_MEMORY;
+	alg->cra_flags = CRYPTO_ALG_ASYNC;
 
 	t_alg->alg_type = template->alg_type;