From: "Alex Bennée" <alex.bennee@linaro.org>
To: Richard Henderson <richard.henderson@linaro.org>
Cc: qemu-devel@nongnu.org
Subject: Re: [PATCH 13/25] tests/tcg: fix constraints in test-i386-adcox
Date: Fri, 28 Feb 2025 11:35:26 +0000 [thread overview]
Message-ID: <87y0xquwsh.fsf@draig.linaro.org> (raw)
In-Reply-To: <5facb17d-c559-46df-be17-b37d22a212bc@linaro.org> (Richard Henderson's message of "Wed, 26 Feb 2025 09:26:24 -0800")
Richard Henderson <richard.henderson@linaro.org> writes:
> On 2/26/25 06:03, Alex Bennée wrote:
>> Clang complains:
>> clang -O2 -m64 -mcx16
>> /home/alex/lsrc/qemu.git/tests/tcg/i386/test-i386-adcox.c -o
>> test-i386-adcox -static
>> /home/alex/lsrc/qemu.git/tests/tcg/i386/test-i386-adcox.c:32:26: error: invalid input constraint '0' in asm
>> : "r" ((REG)-1), "0" (flags), "1" (out_adcx), "2" (out_adox));
>> ^
>> /home/alex/lsrc/qemu.git/tests/tcg/i386/test-i386-adcox.c:57:26: error: invalid input constraint '0' in asm
>> : "r" ((REG)-1), "0" (flags), "1" (out_adcx), "2" (out_adox));
>> ^
>> 2 errors generated.
>> Pointing out a numbered input constraint can't point to a read/write
>> output [1]. Convert to a read-only input constraint to allow this.
>> [1]
>> https://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20101101/036036.html
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
>> ---
>> tests/tcg/i386/test-i386-adcox.c | 4 ++--
>> 1 file changed, 2 insertions(+), 2 deletions(-)
>> diff --git a/tests/tcg/i386/test-i386-adcox.c
>> b/tests/tcg/i386/test-i386-adcox.c
>> index 16169efff8..b56dbc00f2 100644
>> --- a/tests/tcg/i386/test-i386-adcox.c
>> +++ b/tests/tcg/i386/test-i386-adcox.c
>> @@ -28,7 +28,7 @@ void test_adox_adcx(uint32_t in_c, uint32_t in_o, REG adcx_operand, REG adox_ope
>> "adox %3, %2;"
>> "adcx %3, %1;"
>> "pushf; pop %0"
>> - : "+r" (flags), "+r" (out_adcx), "+r" (out_adox)
>> + : "=r"(flags), "=r"(out_adcx), "=r"(out_adox)
>> : "r" ((REG)-1), "0" (flags), "1" (out_adcx), "2" (out_adox));
>
> Alternately, drop all of the matching constraints as redundant.
> That's what "+r" does, anyway.
You mean like this:
asm("push %0; popf;"
"adox %3, %2;"
"adcx %3, %1;"
"pushf; pop %0"
: "+r"(flags), "+r"(out_adcx), "+r"(out_adox));
Because that breaks the test:
test-i386-adcox: /home/alex/lsrc/qemu.git/tests/tcg/i386/test-i386-adcox.c:33: test_adox_adcx: Assertion `out_adcx == in_c + adcx_operand - 1' failed.
>
>
> r~
--
Alex Bennée
Virtualisation Tech Lead @ Linaro
next prev parent reply other threads:[~2025-02-28 11:36 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-26 14:03 [PATCH 00/25] maintainer updates for 10.0 softfreeze (gpu/tcg tests, plugins, MAINTAINERS) Alex Bennée
2025-02-26 14:03 ` [PATCH 01/25] tests/functional: move aarch64 GPU test into own file Alex Bennée
2025-02-26 15:27 ` Thomas Huth
2025-02-26 17:19 ` Richard Henderson
2025-02-26 18:08 ` Alex Bennée
2025-02-26 14:03 ` [PATCH 02/25] tests/functional: factor out common code in gpu test Alex Bennée
2025-02-26 15:32 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 03/25] tests/functional: ensure we have a GPU device for tests Alex Bennée
2025-02-26 15:59 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 04/25] tests/functional: expand tests to cover virgl Alex Bennée
2025-02-26 16:03 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 05/25] plugins: add explicit dependency in functional tests Alex Bennée
2025-02-26 14:03 ` [PATCH 06/25] tests/functional: Introduce the dso_suffix() helper Alex Bennée
2025-02-26 14:03 ` [PATCH 07/25] tests/functional: Allow running TCG plugins tests on non-Linux/BSD hosts Alex Bennée
2025-02-27 9:03 ` Daniel P. Berrangé
2025-02-26 14:03 ` [PATCH 08/25] tests/functional: skip vulkan tests with nVidia Alex Bennée
2025-02-26 16:07 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 09/25] libvirt-ci: bump to latest for vulkan-tools (libvirt MR 525) Alex Bennée
2025-02-26 16:14 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 10/25] tests/vm: bump timeout for shutdown Alex Bennée
2025-02-26 16:15 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 11/25] tests/tcg: mark test-vma as a linux-only test Alex Bennée
2025-02-26 16:16 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 12/25] tests/tcg: add message to _Static_assert in test-avx Alex Bennée
2025-02-26 16:17 ` Thomas Huth
2025-02-26 14:03 ` [PATCH 13/25] tests/tcg: fix constraints in test-i386-adcox Alex Bennée
2025-02-26 17:26 ` Richard Henderson
2025-02-28 11:35 ` Alex Bennée [this message]
2025-02-28 17:35 ` Richard Henderson
2025-02-26 14:03 ` [PATCH 14/25] tests/tcg: enable -fwrapv for test-i386-bmi Alex Bennée
2025-02-26 17:27 ` Richard Henderson
2025-02-26 14:03 ` [PATCH 15/25] plugins/api: use qemu_target_page_mask() to get value Alex Bennée
2025-02-26 17:27 ` Richard Henderson
2025-02-26 22:19 ` Philippe Mathieu-Daudé
2025-02-26 14:03 ` [PATCH 16/25] plugins/loader: populate target_name with target_name() Alex Bennée
2025-02-26 14:03 ` [PATCH 17/25] include/qemu: plugin-memory.h doesn't need cpu-defs.h Alex Bennée
2025-02-26 14:03 ` [PATCH 18/25] plugins/api: clean-up the includes Alex Bennée
2025-02-26 14:03 ` [PATCH 19/25] plugins/plugin.h: include queue.h Alex Bennée
2025-02-26 14:03 ` [PATCH 20/25] plugins/loader: compile loader only once Alex Bennée
2025-02-26 14:03 ` [PATCH 21/25] plugins/api: split out binary path/start/end/entry code Alex Bennée
2025-02-26 15:28 ` Warner Losh
2025-02-26 17:29 ` Richard Henderson
2025-02-26 14:03 ` [PATCH 22/25] plugins/api: split out the vaddr/hwaddr helpers Alex Bennée
2025-02-26 14:03 ` [PATCH 23/25] plugins/api: split out time control helpers Alex Bennée
2025-02-26 14:03 ` [PATCH 24/25] plugins/api: build only once Alex Bennée
2025-02-26 14:03 ` [PATCH 25/25] MAINTAINERS: remove widely sanctioned entities Alex Bennée
2025-02-27 15:45 ` Daniel Henrique Barboza
2025-02-28 10:39 ` Alex Bennée
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=87y0xquwsh.fsf@draig.linaro.org \
--to=alex.bennee@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.