On Fri, Jun 07, 2019 at 09:28:49AM -0500, Richard Henderson wrote: > On 6/7/19 9:09 AM, Laurent Vivier wrote: > > On 07/06/2019 11:29, Laurent Vivier wrote: > >> On 29/05/2019 08:49, David Gibson wrote: > >>> From: Richard Henderson <richard.hender...@linaro.org> > >>> > >>> The gvec expanders take care of masking the shift amount > >>> against the element width. > >>> > >>> Signed-off-by: Richard Henderson <richard.hender...@linaro.org> > >>> Message-Id: <20190518191430.21686-2-richard.hender...@linaro.org> > >>> Signed-off-by: David Gibson <da...@gibson.dropbear.id.au> > >>> --- > >>> target/ppc/helper.h | 12 ---------- > >>> target/ppc/int_helper.c | 37 ----------------------------- > >>> target/ppc/translate/vmx-impl.inc.c | 24 +++++++++---------- > >>> 3 files changed, 12 insertions(+), 61 deletions(-) > >> > >> This patch introduces a regressions > >> with Fedora 29 guest: > >> > >> - during kernel boot: > >> > >> [ 40.397876] crypto_register_alg 'aes' = 0 > >> [ 40.577517] crypto_register_alg 'cbc(aes)' = 0 > >> [ 40.743576] crypto_register_alg 'ctr(aes)' = 0 > >> [ 41.061379] alg: skcipher: Test 1 failed (invalid result) on encryption > >> for p8_aes_xts > >> [ 41.062054] 00000000: 91 7c f6 9e bd 68 b2 ec 9b 9f e9 a3 ea dd a6 92 > >> [ 41.062163] 00000010: 98 10 35 57 5e dc 36 1e 9a f7 bc ba 39 f2 5c eb > >> [ 41.062834] crypto_register_alg 'xts(aes)' = 0 > >> [ 41.077358] alg: hash: Test 2 failed for p8_ghash > >> [ 41.077553] 00000000: 5f 89 ab f7 20 57 20 57 20 57 20 57 20 57 20 57 > >> > >> - with libssl: > >> > >> # curl -o /dev/null https://www.google.com > >> % Total % Received % Xferd Average Speed Time Time Time > >> Current > >> Dload Upload Total Spent Left > >> Speed > >> 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- > >> 0 > >> curl: (35) error:1408F119:SSL routines:ssl3_get_record:decryption failed > >> or bad record mac > >> > >> [before, this one fails with: > >> curl: (35) error:04091068:rsa routines:int_rsa_verify:bad signature ] > >> > >> If I revert this patch on top of 0d74f3b427 + "target/ppc: Fix lxvw4x, > >> lxvh8x and lxvb16x", all works fine. > >> > >> Thanks, > >> Laurent > >> > > > > This seems to fix the problem: > > > > diff --git a/accel/tcg/tcg-runtime-gvec.c b/accel/tcg/tcg-runtime-gvec.c > > index 3b6052fe97..6f0709b307 100644 > > --- a/accel/tcg/tcg-runtime-gvec.c > > +++ b/accel/tcg/tcg-runtime-gvec.c > > @@ -874,7 +874,7 @@ void HELPER(gvec_sar8v)(void *d, void *a, void *b, > > uint32_t desc) > > intptr_t oprsz = simd_oprsz(desc); > > intptr_t i; > > > > - for (i = 0; i < oprsz; i += sizeof(vec8)) { > > + for (i = 0; i < oprsz; i += sizeof(uint8_t)) { > > uint8_t sh = *(uint8_t *)(b + i) & 7; > > *(int8_t *)(d + i) = *(int8_t *)(a + i) >> sh; > > } > > Grr. I really really need to come up with a solution for testing that allows > me to test paths that the host cpu would not ordinarily take. This bug is > hidden on a host with AVX2. > > Thanks for the digging.
Can one of you send this fix formally with a S-o-b and so forth?
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
signature.asc
Description: PGP signature
