From: David Hildenbrand <david@redhat.com>
To: "Bill Wendling" <morbo@google.com>,
kvm@vger.kernel.org, "Thomas Huth" <thuth@redhat.com>,
"Lukáš Doktor" <ldoktor@redhat.com>
Subject: Re: [kvm-unit-tests PATCH] lib: use an argument which doesn't require default argument promotion
Date: Fri, 11 Oct 2019 16:19:07 +0200 [thread overview]
Message-ID: <986a6fc2-ef7b-4df4-8d4e-a4ab94238b32@redhat.com> (raw)
In-Reply-To: <CAGG=3QUL_OrjaWn+gF4z-R8brR2=3661hGk0uUAK2y8Dff7Mvg@mail.gmail.com>
On 10.09.19 01:11, Bill Wendling wrote:
> Clang warns that passing an object that undergoes default argument
> promotion to "va_start" is undefined behavior:
>
> lib/report.c:106:15: error: passing an object that undergoes default
> argument promotion to 'va_start' has undefined behavior
> [-Werror,-Wvarargs]
> va_start(va, pass);
>
> Using an "unsigned" type removes the need for argument promotion.
>
> Signed-off-by: Bill Wendling <morbo@google.com>
> ---
> lib/libcflat.h | 4 ++--
> lib/report.c | 6 +++---
> 2 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/lib/libcflat.h b/lib/libcflat.h
> index b94d0ac..b6635d9 100644
> --- a/lib/libcflat.h
> +++ b/lib/libcflat.h
> @@ -99,9 +99,9 @@ void report_prefix_pushf(const char *prefix_fmt, ...)
> __attribute__((format(printf, 1, 2)));
> extern void report_prefix_push(const char *prefix);
> extern void report_prefix_pop(void);
> -extern void report(const char *msg_fmt, bool pass, ...)
> +extern void report(const char *msg_fmt, unsigned pass, ...)
> __attribute__((format(printf, 1, 3)));
> -extern void report_xfail(const char *msg_fmt, bool xfail, bool pass, ...)
> +extern void report_xfail(const char *msg_fmt, bool xfail, unsigned pass, ...)
> __attribute__((format(printf, 1, 4)));
> extern void report_abort(const char *msg_fmt, ...)
> __attribute__((format(printf, 1, 2)))
> diff --git a/lib/report.c b/lib/report.c
> index ca9b4fd..7d259f6 100644
> --- a/lib/report.c
> +++ b/lib/report.c
> @@ -81,7 +81,7 @@ void report_prefix_pop(void)
> }
>
> static void va_report(const char *msg_fmt,
> - bool pass, bool xfail, bool skip, va_list va)
> + unsigned pass, bool xfail, bool skip, va_list va)
> {
> const char *prefix = skip ? "SKIP"
> : xfail ? (pass ? "XPASS" : "XFAIL")
> @@ -104,7 +104,7 @@ static void va_report(const char *msg_fmt,
> spin_unlock(&lock);
> }
>
> -void report(const char *msg_fmt, bool pass, ...)
> +void report(const char *msg_fmt, unsigned pass, ...)
> {
> va_list va;
> va_start(va, pass);
> @@ -112,7 +112,7 @@ void report(const char *msg_fmt, bool pass, ...)
> va_end(va);
> }
>
> -void report_xfail(const char *msg_fmt, bool xfail, bool pass, ...)
> +void report_xfail(const char *msg_fmt, bool xfail, unsigned pass, ...)
> {
> va_list va;
> va_start(va, pass);
>
This patch breaks the selftest on s390x:
t460s: ~/git/kvm-unit-tests (kein Branch, Rebase von Branch master im Gange) $ ./run_tests.sh
FAIL selftest-setup (14 tests, 2 unexpected failures)
t460s: ~/git/kvm-unit-tests (kein Branch, Rebase von Branch master im Gange) $ cat logs/selftest-setup.log
timeout -k 1s --foreground 90s /usr/bin/qemu-system-s390x -nodefaults -nographic -machine s390-ccw-virtio,accel=tcg -chardev stdio,id=con0 -device sclpconsole,chardev=con0 -kernel s390x/selftest.elf -smp 1 -append test 123 # -initrd /tmp/tmp.JwIjS9RWlv
PASS: selftest: true
PASS: selftest: argc == 3
PASS: selftest: argv[0] == PROGNAME
PASS: selftest: argv[1] == test
PASS: selftest: argv[2] == 123
PASS: selftest: 3.0/2.0 == 1.5
PASS: selftest: Program interrupt: expected(1) == received(1)
PASS: selftest: Program interrupt: expected(5) == received(5)
FAIL: selftest: malloc: got vaddr
PASS: selftest: malloc: access works
FAIL: selftest: malloc: got 2nd vaddr
PASS: selftest: malloc: access works
PASS: selftest: malloc: addresses differ
PASS: selftest: Program interrupt: expected(5) == received(5)
SUMMARY: 14 tests, 2 unexpected failures
EXIT: STATUS=3
A fix for the test would look like this:
diff --git a/s390x/selftest.c b/s390x/selftest.c
index f4acdc4..dc1c476 100644
--- a/s390x/selftest.c
+++ b/s390x/selftest.c
@@ -49,9 +49,10 @@ static void test_malloc(void)
*tmp2 = 123456789;
mb();
- report("malloc: got vaddr", (uintptr_t)tmp & 0xf000000000000000ul);
+ report("malloc: got vaddr", !!((uintptr_t)tmp & 0xf000000000000000ul));
report("malloc: access works", *tmp == 123456789);
- report("malloc: got 2nd vaddr", (uintptr_t)tmp2 & 0xf000000000000000ul);
+ report("malloc: got 2nd vaddr",
+ !!((uintptr_t)tmp2 & 0xf000000000000000ul));
report("malloc: access works", (*tmp2 == 123456789));
report("malloc: addresses differ", tmp != tmp2);
But I am not sure if that is the right fix.
(why don't we run sanity tests to detect that, this tests works
just fine with s390x TCG)
--
Thanks,
David / dhildenb
next prev parent reply other threads:[~2019-10-11 14:19 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-09 23:11 [kvm-unit-tests PATCH] lib: use an argument which doesn't require default argument promotion Bill Wendling
2019-09-10 16:43 ` Jim Mattson
2019-09-30 21:59 ` Bill Wendling
2019-10-09 10:13 ` Paolo Bonzini
2019-10-11 14:19 ` David Hildenbrand [this message]
2019-10-11 16:24 ` Thomas Huth
2019-10-11 16:36 ` Thomas Huth
2019-10-11 18:36 ` Bill Wendling
2019-10-14 7:57 ` Thomas Huth
2019-10-14 8:12 ` David Hildenbrand
2019-10-14 8:14 ` Bill Wendling
2019-10-14 9:32 ` Thomas Huth
2019-10-14 11:04 ` Paolo Bonzini
2019-10-17 12:32 ` Thomas Huth
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=986a6fc2-ef7b-4df4-8d4e-a4ab94238b32@redhat.com \
--to=david@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=ldoktor@redhat.com \
--cc=morbo@google.com \
--cc=thuth@redhat.com \
/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).