Re: Call for testers: FPU changes
--o+NaKOy/ZgNfbXd2
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Mon, Nov 15, 2010 at 10:42:50PM -0500, Mike Tancsa wrote:
> On 11/15/2010 4:13 PM, Kostik Belousov wrote:
> >=20
> > Patch is at
> > http://people.freebsd.org/~kib/misc/releng_8_fpu.1.patch
>=20
>=20
> Hi,
> One small failure on the patch
>=20
> The text leading up to this was:
> --------------------------
> |Index: pc98/include/npx.h
> |=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> |--- pc98/include/npx.h (revision 215253)
> |+++ pc98/include/npx.h (working copy)
> --------------------------
> Patching file pc98/include/npx.h using Plan A...
> Hunk #1 failed at 1.
> 1 out of 1 hunks failed--saving rejects to pc98/include/npx.h.rej
This is because our patch(1) in base is somewhat old, I believe.
The diff was generated by svn diff from the up to date stable/8
checkout, and the reason for failure is expanded $FreeBSD$ tags.
Newer gnu patch, available in ports, handless this correctly,
reporting about patches applied with "fuzz".
>=20
>=20
> I tested with openssl and openvpn and all seems to work great on the via
> board and my i5 board!! Simple test details at
>=20
> http://www.tancsa.com/fpu.html
>=20
> I will try out geli and some more extensive tests tomorrow
>=20
> Thanks for porting this back to RELENG_8 !
This is actually somewhat puzzling. Does openssl in base automatically
use crypto(4) ?
Also, could you, please redo the speed tests for aesni(4) with the
following patch applied over the driver sources ?
Thank you !
diff --git a/sys/crypto/aesni/aesni_wrap.c b/sys/crypto/aesni/aesni_wrap.c
index 36c66ea..3fd397c 100644
--- a/sys/crypto/aesni/aesni_wrap.c
+++ b/sys/crypto/aesni/aesni_wrap.c
@@ -246,14 +246,21 @@ int
aesni_cipher_setup(struct aesni_session *ses, struct cryptoini *encini)
{
struct thread *td;
- int error;
+ int error, saved_ctx;
=20
td =3D curthread;
- error =3D fpu_kern_enter(td, &ses->fpu_ctx, FPU_KERN_NORMAL);
+ if (!is_fpu_kern_thread(0)) {
+ error =3D fpu_kern_enter(td, &ses->fpu_ctx, FPU_KERN_NORMAL);
+ saved_ctx =3D 1;
+ } else {
+ error =3D 0;
+ saved_ctx =3D 0;
+ }
if (error =3D=3D 0) {
error =3D aesni_cipher_setup_common(ses, encini->cri_key,
encini->cri_klen);
- fpu_kern_leave(td, &ses->fpu_ctx);
+ if (saved_ctx)
+ fpu_kern_leave(td, &ses->fpu_ctx);
}
return (error);
}
@@ -264,16 +271,22 @@ aesni_cipher_process(struct aesni_session *ses, struc=
t cryptodesc *enccrd,
{
struct thread *td;
uint8_t *buf;
- int error, allocated;
+ int error, allocated, saved_ctx;
=20
buf =3D aesni_cipher_alloc(enccrd, crp, &allocated);
if (buf =3D=3D NULL)
return (ENOMEM);
=20
td =3D curthread;
- error =3D fpu_kern_enter(td, &ses->fpu_ctx, FPU_KERN_NORMAL);
- if (error !=3D 0)
- goto out;
+ if (!is_fpu_kern_thread(0)) {
+ error =3D fpu_kern_enter(td, &ses->fpu_ctx, FPU_KERN_NORMAL);
+ if (error !=3D 0)
+ goto out;
+ saved_ctx =3D 1;
+ } else {
+ saved_ctx =3D 0;
+ error =3D 0;
+ }
=20
if ((enccrd->crd_flags & CRD_F_KEY_EXPLICIT) !=3D 0) {
error =3D aesni_cipher_setup_common(ses, enccrd->crd_key,
@@ -311,7 +324,8 @@ aesni_cipher_process(struct aesni_session *ses, struct =
cryptodesc *enccrd,
ses->iv);
}
}
- fpu_kern_leave(td, &ses->fpu_ctx);
+ if (saved_ctx)
+ fpu_kern_leave(td, &ses->fpu_ctx);
if (allocated)
crypto_copyback(crp->crp_flags, crp->crp_buf, enccrd->crd_skip,
enccrd->crd_len, buf);
--o+NaKOy/ZgNfbXd2
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (FreeBSD)
iEYEARECAAYFAkziUkIACgkQC3+MBN1Mb4iI5QCdHlqsjEY2d4IPan9+5EcabUXb
0nIAoMyrnETGquYpttWzQ7lwQhAySMsL
=T1Bt
-----END PGP SIGNATURE-----
--o+NaKOy/ZgNfbXd2--
討論串 (同標題文章)
完整討論串 (本文為第 3 之 13 篇)：