* [PATCH] Fix subtle race in tst-cancel2 / tst-cancelx2
@ 2017-12-08 17:47 Vineet Gupta
2017-12-08 18:07 ` [PATCH v2] " Vineet Gupta
0 siblings, 1 reply; 4+ messages in thread
From: Vineet Gupta @ 2017-12-08 17:47 UTC (permalink / raw)
To: linux-snps-arc
When ran on ARC, these tests would ocassionally fail
| [ARCLinux]# for i in 1 2 3 4 5 ; do ./tst-cancel2; echo $?; done
| write succeeded
| result is wrong: expected 0xffffffff, got 0x1
| 1 <-- fail
| 0 <-- pass
| 0 <--- pass
| 0 <-- pass
| write succeeded
| result is wrong: expected 0xffffffff, got 0x1
| 1 <-- fail
Same test (which originated form glibc) doesn't fail in glibc builds.
Turns out there's a subtle race in uclibc version
The test creates a new thread, makes it do a looong write call, and
parent then cancels the thread, expecting it to unwind out of write
call cleanly. However the write (even for 10k bytes) could fnish befor
eparent gets a chance ti run and/or cancel - causing the occasional
failire.
Cc: Cupertino Miranda <cmiranda at synopsys.com>
Fix this subtelty by making it write not just once but forever.
Signed-off-by: Vineet Gupta <vgupta at synopsys.com>
---
test/nptl/tst-cancel2.c | 6 +-----
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/test/nptl/tst-cancel2.c b/test/nptl/tst-cancel2.c
index 45c9e8ea957a..08dd13b10f37 100644
--- a/test/nptl/tst-cancel2.c
+++ b/test/nptl/tst-cancel2.c
@@ -32,11 +32,7 @@ tf (void *arg)
write blocks. */
char buf[100000];
- if (write (fd[1], buf, sizeof (buf)) == sizeof (buf))
- {
- puts ("write succeeded");
- return (void *) 1l;
- }
+ while (write (fd[1], buf, sizeof (buf)) > 0);
return (void *) 42l;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH v2] Fix subtle race in tst-cancel2 / tst-cancelx2
2017-12-08 17:47 [PATCH] Fix subtle race in tst-cancel2 / tst-cancelx2 Vineet Gupta
@ 2017-12-08 18:07 ` Vineet Gupta
2017-12-18 19:18 ` [uclibc-ng-devel] " Vineet Gupta
0 siblings, 1 reply; 4+ messages in thread
From: Vineet Gupta @ 2017-12-08 18:07 UTC (permalink / raw)
To: linux-snps-arc
When ran on ARC, these tests would ocassionally fail
| [ARCLinux]# for i in 1 2 3 4 5 ; do ./tst-cancel2; echo $?; done
| write succeeded
| result is wrong: expected 0xffffffff, got 0x1
| 1 <-- fail
| 0 <-- pass
| 0 <--- pass
| 0 <-- pass
| write succeeded
| result is wrong: expected 0xffffffff, got 0x1
| 1 <-- fail
Same test (which originated form glibc) doesn't fail in glibc builds.
Turns out there's a subtle race in uclibc version
The test creates a new thread, makes it do a looong write call, and
parent then cancels the thread, expecting it to unwind out of write
call cleanly. However the write (even for 10k bytes) could finish
before parent gets a chance to resume and/or cancel it, causing the
occasional failure.
Fix this subtelty by making it write not just once but forever.
Cc: Cupertino Miranda <cmiranda at synopsys.com>
Signed-off-by: Vineet Gupta <vgupta at synopsys.com>
---
Change since v1: fix typos in changelogs
---
test/nptl/tst-cancel2.c | 6 +-----
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/test/nptl/tst-cancel2.c b/test/nptl/tst-cancel2.c
index 45c9e8ea957a..08dd13b10f37 100644
--- a/test/nptl/tst-cancel2.c
+++ b/test/nptl/tst-cancel2.c
@@ -32,11 +32,7 @@ tf (void *arg)
write blocks. */
char buf[100000];
- if (write (fd[1], buf, sizeof (buf)) == sizeof (buf))
- {
- puts ("write succeeded");
- return (void *) 1l;
- }
+ while (write (fd[1], buf, sizeof (buf)) > 0);
return (void *) 42l;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [uclibc-ng-devel] [PATCH v2] Fix subtle race in tst-cancel2 / tst-cancelx2
2017-12-08 18:07 ` [PATCH v2] " Vineet Gupta
@ 2017-12-18 19:18 ` Vineet Gupta
2017-12-18 19:40 ` Waldemar Brodkorb
0 siblings, 1 reply; 4+ messages in thread
From: Vineet Gupta @ 2017-12-18 19:18 UTC (permalink / raw)
To: linux-snps-arc
On 12/08/2017 10:07 AM, Vineet Gupta wrote:
> When ran on ARC, these tests would ocassionally fail
>
> | [ARCLinux]# for i in 1 2 3 4 5 ; do ./tst-cancel2; echo $?; done
> | write succeeded
> | result is wrong: expected 0xffffffff, got 0x1
> | 1 <-- fail
> | 0 <-- pass
> | 0 <--- pass
> | 0 <-- pass
> | write succeeded
> | result is wrong: expected 0xffffffff, got 0x1
> | 1 <-- fail
>
> Same test (which originated form glibc) doesn't fail in glibc builds.
> Turns out there's a subtle race in uclibc version
>
> The test creates a new thread, makes it do a looong write call, and
> parent then cancels the thread, expecting it to unwind out of write
> call cleanly. However the write (even for 10k bytes) could finish
> before parent gets a chance to resume and/or cancel it, causing the
> occasional failure.
>
> Fix this subtelty by making it write not just once but forever.
>
> Cc: Cupertino Miranda <cmiranda at synopsys.com>
> Signed-off-by: Vineet Gupta <vgupta at synopsys.com>
> ---
> Change since v1: fix typos in changelogs
> ---
Ping ?
> test/nptl/tst-cancel2.c | 6 +-----
> 1 file changed, 1 insertion(+), 5 deletions(-)
>
> diff --git a/test/nptl/tst-cancel2.c b/test/nptl/tst-cancel2.c
> index 45c9e8ea957a..08dd13b10f37 100644
> --- a/test/nptl/tst-cancel2.c
> +++ b/test/nptl/tst-cancel2.c
> @@ -32,11 +32,7 @@ tf (void *arg)
> write blocks. */
> char buf[100000];
>
> - if (write (fd[1], buf, sizeof (buf)) == sizeof (buf))
> - {
> - puts ("write succeeded");
> - return (void *) 1l;
> - }
> + while (write (fd[1], buf, sizeof (buf)) > 0);
>
> return (void *) 42l;
> }
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* [uclibc-ng-devel] [PATCH v2] Fix subtle race in tst-cancel2 / tst-cancelx2
2017-12-18 19:18 ` [uclibc-ng-devel] " Vineet Gupta
@ 2017-12-18 19:40 ` Waldemar Brodkorb
0 siblings, 0 replies; 4+ messages in thread
From: Waldemar Brodkorb @ 2017-12-18 19:40 UTC (permalink / raw)
To: linux-snps-arc
Hi,
patch applied and pushed,
just forgot the mail,
best regards
Waldemar
> Am 18.12.2017 um 20:18 schrieb Vineet Gupta <Vineet.Gupta1 at synopsys.com>:
>
>> On 12/08/2017 10:07 AM, Vineet Gupta wrote:
>> When ran on ARC, these tests would ocassionally fail
>> | [ARCLinux]# for i in 1 2 3 4 5 ; do ./tst-cancel2; echo $?; done
>> | write succeeded
>> | result is wrong: expected 0xffffffff, got 0x1
>> | 1 <-- fail
>> | 0 <-- pass
>> | 0 <--- pass
>> | 0 <-- pass
>> | write succeeded
>> | result is wrong: expected 0xffffffff, got 0x1
>> | 1 <-- fail
>> Same test (which originated form glibc) doesn't fail in glibc builds.
>> Turns out there's a subtle race in uclibc version
>> The test creates a new thread, makes it do a looong write call, and
>> parent then cancels the thread, expecting it to unwind out of write
>> call cleanly. However the write (even for 10k bytes) could finish
>> before parent gets a chance to resume and/or cancel it, causing the
>> occasional failure.
>> Fix this subtelty by making it write not just once but forever.
>> Cc: Cupertino Miranda <cmiranda at synopsys.com>
>> Signed-off-by: Vineet Gupta <vgupta at synopsys.com>
>> ---
>> Change since v1: fix typos in changelogs
>> ---
>
> Ping ?
>
>
>> test/nptl/tst-cancel2.c | 6 +-----
>> 1 file changed, 1 insertion(+), 5 deletions(-)
>> diff --git a/test/nptl/tst-cancel2.c b/test/nptl/tst-cancel2.c
>> index 45c9e8ea957a..08dd13b10f37 100644
>> --- a/test/nptl/tst-cancel2.c
>> +++ b/test/nptl/tst-cancel2.c
>> @@ -32,11 +32,7 @@ tf (void *arg)
>> write blocks. */
>> char buf[100000];
>> - if (write (fd[1], buf, sizeof (buf)) == sizeof (buf))
>> - {
>> - puts ("write succeeded");
>> - return (void *) 1l;
>> - }
>> + while (write (fd[1], buf, sizeof (buf)) > 0);
>> return (void *) 42l;
>> }
>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2017-12-18 19:40 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-12-08 17:47 [PATCH] Fix subtle race in tst-cancel2 / tst-cancelx2 Vineet Gupta
2017-12-08 18:07 ` [PATCH v2] " Vineet Gupta
2017-12-18 19:18 ` [uclibc-ng-devel] " Vineet Gupta
2017-12-18 19:40 ` Waldemar Brodkorb
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.