Mailing List Archive

[PATCH] cipher/sha512: Fix non-NEON ARM assembly implementation
* cipher/sha512.c (do_transform_generic)
[USE_ARM_ASM]: Switch to the non-NEON assembly implementation.

--

When building for ARM CPUs that don't support NEON, linking fails
with an "undefined reference to _gcry_sha512_transform_armv7_neon"
error. Switching to the non-NEON assembly function corrects this.
---
cipher/sha512.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cipher/sha512.c b/cipher/sha512.c
index f70cdf42..0f4c304f 100644
--- a/cipher/sha512.c
+++ b/cipher/sha512.c
@@ -291,7 +291,7 @@ static unsigned int
do_transform_generic (void *context, const unsigned char *data, size_t nblks)
{
SHA512_CONTEXT *hd = context;
- return _gcry_sha512_transform_armv7_neon (&hd->state, data, k, nblks);
+ return _gcry_sha512_transform_arm (&hd->state, data, k, nblks);
}
#else
static unsigned int
--
2.26.2

_______________________________________________
Gnupg-devel mailing list
Gnupg-devel@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-devel
Re: [PATCH] cipher/sha512: Fix non-NEON ARM assembly implementation [ In reply to ]
Hi!

Thanks David for the reports. I accidentally mentioned this mailing
list in the annoucnement, but it really this should go to gcrypt-devel.
Libgcrypt hackers are not necessary reading this list.
I'll repost your mails to gcrypt-devel.

We have a couple of fixes already in the repo, see also
https://dev/gnupg.org/T5251 for the NEON thing. I am not sure whether
the Poly1305 bug has already been reported.




Shalom-Salam,

Werner


--
* Free Assange and protect free journalism!
* Germany: Sign the Treaty on the Prohibition of Nuclear Weapons!