@@ -760,14 +760,15 @@ static int gcmaes_crypt_by_sg(bool enc, struct aead_request *req,
gcm_tfm = &aesni_gcm_tfm_sse;
/* Linearize assoc, if not already linear */
- if (req->src->length >= assoclen && req->src->length &&
- (!PageHighMem(sg_page(req->src)) ||
- req->src->offset + req->src->length <= PAGE_SIZE)) {
+ if (req->src->length >= assoclen && req->src->length) {
scatterwalk_start(&assoc_sg_walk, req->src);
assoc = scatterwalk_map(&assoc_sg_walk);
} else {
+ gfp_t flags = (req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP)
+ ? GFP_KERNEL : GFP_ATOMIC;
+
/* assoc can be any length, so must be on heap */
- assocmem = kmalloc(assoclen, GFP_ATOMIC);
+ assocmem = kmalloc(assoclen, flags);
if (unlikely(!assocmem))
return -ENOMEM;
assoc = assocmem;
The gcm(aes-ni) driver is only built for x86_64, which does not make use of highmem. So testing for PageHighMem is pointless and can be omitted. While at it, replace GFP_ATOMIC with the appropriate runtime decided value based on the context. Signed-off-by: Ard Biesheuvel <ardb@kernel.org> --- arch/x86/crypto/aesni-intel_glue.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)