From: Aurelien Jarno <aurelien@aurel32.net>
To: Tristan Gingold <gingold@adacore.com>
Cc: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH] Alpha bug fix: overflow condition for sublv and subqv
Date: Fri, 10 Apr 2009 23:26:35 +0200 [thread overview]
Message-ID: <20090410212635.GA8217@volta.aurel32.net> (raw)
In-Reply-To: <1239353720-67935-1-git-send-email-gingold@adacore.com>
On Fri, Apr 10, 2009 at 10:55:20AM +0200, Tristan Gingold wrote:
> The conditions to detect overflow in sub operations was wrong.
>
> This patch is necessary to boot Tru64.
>
> Signed-off-by: Tristan Gingold <gingold@adacore.com>
> ---
> target-alpha/op_helper.c | 20 ++++++++++----------
> tests/alpha/Makefile | 3 +++
> tests/alpha/test-ovf.c | 29 +++++++++++++++++++++++++++++
> 3 files changed, 42 insertions(+), 10 deletions(-)
> create mode 100644 tests/alpha/test-ovf.c
Thanks, applied.
> diff --git a/target-alpha/op_helper.c b/target-alpha/op_helper.c
> index 7ad1b3d..3c95e98 100644
> --- a/target-alpha/op_helper.c
> +++ b/target-alpha/op_helper.c
> @@ -158,22 +158,22 @@ uint64_t helper_addlv (uint64_t op1, uint64_t op2)
>
> uint64_t helper_subqv (uint64_t op1, uint64_t op2)
> {
> - uint64_t tmp = op1;
> - op1 -= op2;
> - if (unlikely(((~tmp) ^ op1 ^ (-1ULL)) & ((~tmp) ^ op2) & (1ULL << 63))) {
> - helper_excp(EXCP_ARITH, EXCP_ARITH_OVERFLOW);
> + uint64_t res;
> + res = op1 - op2;
> + if (unlikely((op1 ^ op2) & (res ^ op1) & (1ULL << 63))) {
> + helper_excp(EXCP_GEN_ARITH, EXCP_ARITH_OVERFLOW);
> }
> - return op1;
> + return res;
> }
>
> uint64_t helper_sublv (uint64_t op1, uint64_t op2)
> {
> - uint64_t tmp = op1;
> - op1 = (uint32_t)(op1 - op2);
> - if (unlikely(((~tmp) ^ op1 ^ (-1UL)) & ((~tmp) ^ op2) & (1UL << 31))) {
> - helper_excp(EXCP_ARITH, EXCP_ARITH_OVERFLOW);
> + uint32_t res;
> + res = op1 - op2;
> + if (unlikely((op1 ^ op2) & (res ^ op1) & (1UL << 31))) {
> + helper_excp(EXCP_GEN_ARITH, EXCP_ARITH_OVERFLOW);
> }
> - return op1;
> + return res;
> }
>
> uint64_t helper_mullv (uint64_t op1, uint64_t op2)
> diff --git a/tests/alpha/Makefile b/tests/alpha/Makefile
> index e4e23d2..2b1f03d 100644
> --- a/tests/alpha/Makefile
> +++ b/tests/alpha/Makefile
> @@ -23,6 +23,9 @@ test-cmov.o: test-cond.c
> test-cmov: test-cmov.o crt.o
> $(LINK)
>
> +test-ovf: test-ovf.o crt.o
> + $(LINK)
> +
> check: $(TESTS)
> for f in $(TESTS); do $(SIM) $$f || exit 1; done
>
> diff --git a/tests/alpha/test-ovf.c b/tests/alpha/test-ovf.c
> new file mode 100644
> index 0000000..01c80e7
> --- /dev/null
> +++ b/tests/alpha/test-ovf.c
> @@ -0,0 +1,29 @@
> +static long test_subqv (long a, long b)
> +{
> + long res;
> +
> + asm ("subq/v %1,%2,%0"
> + : "=r" (res) : "r" (a), "r" (b));
> + return res;
> +}
> +static struct {
> + long (*func)(long, long);
> + long a;
> + long b;
> + long r;
> +} vectors[] =
> + {
> + {test_subqv, 0, 0x7d54000, 0xfffffffff82ac000L}
> + };
> +
> +int main (void)
> +{
> + int i;
> +
> + for (i = 0; i < sizeof (vectors)/sizeof(vectors[0]); i++)
> + if ((*vectors[i].func)(vectors[i].a, vectors[i].b) != vectors[i].r) {
> + write(1, "Failed\n", 7);
> + }
> + write(1, "OK\n", 3);
> + return 0;
> +}
> --
> 1.6.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
prev parent reply other threads:[~2009-04-10 21:26 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-04-10 8:55 [Qemu-devel] [PATCH] Alpha bug fix: overflow condition for sublv and subqv Tristan Gingold
2009-04-10 21:26 ` Aurelien Jarno [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20090410212635.GA8217@volta.aurel32.net \
--to=aurelien@aurel32.net \
--cc=gingold@adacore.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).