Skip to contentSkip to conversation

crypto: sun8i-ss: fix multiple memory leaks in sun8i_ss_hash_run

PATCH In review Dongliang Mu 2 participants 3 messages Jul 26, 2021 – Aug 2, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v4 1/1 crypto: sun8i-ss: fix multiple memory leaks in sun8i_ss_hash_run

In sun8i_ss_hash_run, all the dma_mmap_sg/single will cause memory leak due to no corresponding unmap operation if errors happen.

Fix this by adding error handling part for all the dma_mmap_sg/single.

Fixes d9b45418a917 ("crypto: sun8i-ss - support hash algorithms")

---
v1->v2: move crypto_finalize_hash_request to the end of function; move the memcpy after the dma_mmap_sg/single functions.
v2->v3: remove some unrelated code changes; delete the fix of return value since there is no corresponding handling code
v3->v4: change sun8i_ce to sun8i_ss

 drivers/crypto/allwinner/sun8i-ss/sun8i-ss-hash.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/crypto/allwinner/sun8i-ss/sun8i-ss-hash.c +8 -7
index 3c073eb3db03..5448705e8ae1 100644
@@ -368,14 +368,14 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
 	if (nr_sgs <= 0 || nr_sgs > MAX_SG) {
 		dev_err(ss->dev, "Invalid sg number %d\n", nr_sgs);
 		err = -EINVAL;
-		goto theend;
+		goto err_result;
 	}
 
 	addr_res = dma_map_single(ss->dev, result, digestsize, DMA_FROM_DEVICE);
 	if (dma_mapping_error(ss->dev, addr_res)) {
 		dev_err(ss->dev, "DMA map dest\n");
 		err = -EINVAL;
-		goto theend;
+		goto err_unmap_sg;
 	}
 
 	len = areq->nbytes;
@@ -390,7 +390,7 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
 	if (len > 0) {
 		dev_err(ss->dev, "remaining len %d\n", len);
 		err = -EINVAL;
-		goto theend;
+		goto err_addr_res;
 	}
 
 	byte_count = areq->nbytes;
@@ -428,18 +428,19 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
 	if (dma_mapping_error(ss->dev, addr_pad)) {
 		dev_err(ss->dev, "DMA error on padding SG\n");
 		err = -EINVAL;
-		goto theend;
+		goto err_addr_res;
 	}
 
 	err = sun8i_ss_run_hash_task(ss, rctx, crypto_tfm_alg_name(areq->base.tfm));
 
 	dma_unmap_single(ss->dev, addr_pad, j * 4, DMA_TO_DEVICE);
+err_addr_res:
+	dma_unmap_single(ss->dev, addr_res, digestsize, DMA_FROM_DEVICE);
+err_unmap_sg:
 	dma_unmap_sg(ss->dev, areq->src, sg_nents(areq->src),
 		     DMA_TO_DEVICE);
-	dma_unmap_single(ss->dev, addr_res, digestsize, DMA_FROM_DEVICE);
-
 	memcpy(areq->result, result, algt->alg.hash.halg.digestsize);
-theend:
+err_result:
 	kfree(pad);
 	kfree(result);
 	crypto_finalize_hash_request(engine, breq, err);
Signature · 1 line
2.25.1
#1 lore raw
Quoted 6 lines · Dongliang Mu

In sun8i_ss_hash_run, all the dma_mmap_sg/single will cause memory leak due to no corresponding unmap operation if errors happen.

Fix this by adding error handling part for all the dma_mmap_sg/single.

Quoted 8 lines
Fixes d9b45418a917 ("crypto: sun8i-ss - support hash algorithms")

---
v1->v2: move crypto_finalize_hash_request to the end of function; move the memcpy after the dma_mmap_sg/single functions.
v2->v3: remove some unrelated code changes; delete the fix of return value since there is no corresponding handling code
v3->v4: change sun8i_ce to sun8i_ss

From v3 to v4, I changed the title and commit message from sun8i_ce to sun8i_ss because I suddenly realized that the code change below is for sun8i_ss_hash_run, other than sun8i_ce_hash_run.

Furthermore, sun8i_ce is also prone to this issue. I've sent another patch to fix it.

Quoted 60 lines
 drivers/crypto/allwinner/sun8i-ss/sun8i-ss-hash.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/crypto/allwinner/sun8i-ss/sun8i-ss-hash.c +8 -8
index 3c073eb3db03..5448705e8ae1 100644
@@ -368,14 +368,14 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
        if (nr_sgs <= 0 || nr_sgs > MAX_SG) {
                dev_err(ss->dev, "Invalid sg number %d\n", nr_sgs);
                err = -EINVAL;
-               goto theend;
+               goto err_result;
        }
 
        addr_res = dma_map_single(ss->dev, result, digestsize, DMA_FROM_DEVICE);
        if (dma_mapping_error(ss->dev, addr_res)) {
                dev_err(ss->dev, "DMA map dest\n");
                err = -EINVAL;
-               goto theend;
+               goto err_unmap_sg;
        }
 
        len = areq->nbytes;
@@ -390,7 +390,7 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
        if (len > 0) {
                dev_err(ss->dev, "remaining len %d\n", len);
                err = -EINVAL;
-               goto theend;
+               goto err_addr_res;
        }
 
        byte_count = areq->nbytes;
@@ -428,18 +428,19 @@ int sun8i_ss_hash_run(struct crypto_engine *engine, void *breq)
        if (dma_mapping_error(ss->dev, addr_pad)) {
                dev_err(ss->dev, "DMA error on padding SG\n");
                err = -EINVAL;
-               goto theend;
+               goto err_addr_res;
        }
 
        err = sun8i_ss_run_hash_task(ss, rctx, crypto_tfm_alg_name(areq->base.tfm));
 
        dma_unmap_single(ss->dev, addr_pad, j * 4, DMA_TO_DEVICE);
+err_addr_res:
+       dma_unmap_single(ss->dev, addr_res, digestsize, DMA_FROM_DEVICE);
+err_unmap_sg:
        dma_unmap_sg(ss->dev, areq->src, sg_nents(areq->src),
                     DMA_TO_DEVICE);
-       dma_unmap_single(ss->dev, addr_res, digestsize, DMA_FROM_DEVICE);
-
        memcpy(areq->result, result, algt->alg.hash.halg.digestsize);
-theend:
+err_result:
        kfree(pad);
        kfree(result);
        crypto_finalize_hash_request(engine, breq, err);
--

2.25.1

maintainer
#2 lore raw
In reply to Dongliang Mu
Quoted 5 lines · Dongliang Mu

In sun8i_ss_hash_run, all the dma_mmap_sg/single will cause memory leak due to no corresponding unmap operation if errors happen.

Fix this by adding error handling part for all the dma_mmap_sg/single.

Hello

I have nearly the same comment than on your sun8i-ce patch, (copy result only if err===0 and wording).

Regards