From: LiSong
There is no guarantee that the LOOP of the child proecess
can be certainly big enough on any device at any time.
So if set_affinity function is available,
use it to bind the test processes to a single CPU
rather than using the LOOP.
Signed-off-by: Song.Li
---
.../conformance/inter
Hi!
> >
> > Hmm, I would make sure here, that I haven't read more than SIZE
> > characters. Eg. I wouldn't expect that device you are testing is null
> > terminated, even if you wrote the terminating zero there.
>
> How about:
>
> while (s[i] && i < SIZE-1) {
> if (s[i] != 'a')
> fail++;
>
On 12/23/2011 12:51 AM, Cyril Hrubis wrote:
> Hi!
>> diff --git a/testcases/kernel/mem/zram/zram01.c
>> b/testcases/kernel/mem/zram/zram01.c
>> index 6967bd1..38adf63 100644
>> --- a/testcases/kernel/mem/zram/zram01.c
>> +++ b/testcases/kernel/mem/zram/zram01.c
>> @@ -44,6 +44,7 @@ int modprobe =
Hi!
> >> Sometimes `if (access(...) && errno == ENOENT)' doesn't work due to
> >> errno not captures correctly from access() exection. This patch put them
> >> into 2 nested if-condition statements.
> >>
> >> Signed-off-by: Caspar Zhang
> >> ---
> >> testcases/kernel/mem/zram/zram01.c | 20
On 12/23/2011 12:44 AM, Cyril Hrubis wrote:
> Hi!
>> Sometimes `if (access(...) && errno == ENOENT)' doesn't work due to
>> errno not captures correctly from access() exection. This patch put them
>> into 2 nested if-condition statements.
>>
>> Signed-off-by: Caspar Zhang
>> ---
>> testcases/kern
Hi!
> diff --git a/testcases/kernel/mem/zram/zram01.c
> b/testcases/kernel/mem/zram/zram01.c
> index 6967bd1..38adf63 100644
> --- a/testcases/kernel/mem/zram/zram01.c
> +++ b/testcases/kernel/mem/zram/zram01.c
> @@ -44,6 +44,7 @@ int modprobe = 0;
>
> static void set_disksize(void);
> static
Hi!
> Sometimes `if (access(...) && errno == ENOENT)' doesn't work due to
> errno not captures correctly from access() exection. This patch put them
> into 2 nested if-condition statements.
>
> Signed-off-by: Caspar Zhang
> ---
> testcases/kernel/mem/zram/zram01.c | 20 ++--
>
Hi!
> test_proc_kill() tests process will be killed due to exceed memory limit, but
> when memcg_process exit due to mmap/munmap fails, the case still pass.
>
> Signed-off-by: Peng Haitao
> ---
> .../memcg/functional/memcg_function_test.sh| 12 +++-
> 1 files changed, 11 insert
On 11/18/2011 05:37 PM, Caspar Zhang wrote:
> Hi Cyril, Wanlong and all,
>
>This is the latest version of mm/zram01 testcase fixes. Current
>zram01 test has mulitple issues:
>
>a) typo error
>b) it fails to give correct behavior if zram not mounted.
>c) it fails to repeat the
On 10/18/2011 07:02 PM, Caspar Zhang wrote:
>
> swapping01 - first time swap use results in heavy swapping
>
> This case is used for testing upstream commit:
> 50a15981a1fac7e019ff7c3cba87531fb580f065
>
> The upstream commit fixed a issue on s390 platform that heavy
> swapping might occur in som
This patch adds getrusage04 test into runtest/syscalls
Signed-off-by: Caspar Zhang
---
runtest/syscalls | 11 ++-
1 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/runtest/syscalls b/runtest/syscalls
index 3755789..d803022 100644
--- a/runtest/syscalls
+++ b/runtest/sysca
Hi!
> Remove the out-of-dated e4defrag prog, and use the e4defrag
> we installed own.
>
> Note:As Kazuya said, e4defrag here is too old to defrag the
> file in the latest kernel. We should use the latest e4defrag
> command that is contained in e2fsprogs. So, when testing,
> install the lastest e2f
Hi!
> check_mem_stat() need only one argument, so the second argument is redundant.
>
> Signed-off-by: Peng Haitao
Commited, thanks.
--
Cyril Hrubis
chru...@suse.cz
--
Write once. Port to many.
Get the SDK and tools t
Hi!
> Import a new semphore semsync2 to sync thread parent and child thread
> 'th_work', so that signal handlers are installed before receive signal
> USR1 and USR2.
Commited, as this is better than the previous state.
However the testcase implementation is quite questionable. It's trying
to hamm
Hi!
> diff --git a/runtest/syscalls b/runtest/syscalls
> index 3056659..48caaae 100644
> --- a/runtest/syscalls
> +++ b/runtest/syscalls
> @@ -420,6 +420,8 @@ getuid02_16 getuid02_16
> getuid03 getuid03
> getuid03_16 getuid03_16
>
> +getxattr01 getxattr01
> +
> #Needs tty device.
> #ioctl01 i
Hi!
> diff --git a/testcases/kernel/syscalls/setxattr/setxattr01.c
> b/testcases/kernel/syscalls/setxattr/setxattr01.c
> new file mode 100644
> index 000..cfe5e52
> --- /dev/null
> +++ b/testcases/kernel/syscalls/setxattr/setxattr01.c
> @@ -0,0 +1,207 @@
> +/*
> + * Copyright (C) 2011 Red Hat,
Hi!
>
> getrusage04 - accuracy of getrusage() with RUSAGE_THREAD
>
> This program is used for testing the following upstream commit:
> 761b1d26df542fd5eb348837351e4d2f3bc7bffe.
>
> getrusage() returns cpu resource usage with accuracy of 10ms
> when RUSAGE_THREAD is specified to the argument who.
Hi!
> while (wait--)
> ;
>
Hmm, I do not fancy the semicolon on next line, but checkpatch doesn't
seem to have problem with that...
Anyway this version seems good.
--
Cyril Hrubis
chru...@suse.cz
--
Write once. Po
On 12/22/2011 03:51 PM, Caspar Zhang wrote:
> On 11/15/2011 09:19 PM, Zhouping Liu wrote:
>> hi, Cyril & Garrett
>>
>> please feel free to comment the patch.
>>
>> There are two tunables overcommit_memory and overcommit_ratio under
>> /proc/sys/vm/, which can control memory overcommitment.
>>
>> Th
Hi!
> On powerpc platform, getrusage03_child returns the following error
> message:
>
> getrusage03_child1 TBROK : OPTION PARSING ERROR - parse_opts: ERROR -
> Only short options are allowed
>
> This is because child_options[] didn't end up with a NULL array. This
> patch fixes this probl
Basic tests for getxattr(2) and make sure getxattr(2) handles error
conditions correctly.
There are 4 test cases:
1. Get an non-existing attribute,
getxattr(2) should return -1 and set errno to ENOATTR
2. Buffer size is smaller than attribute value size,
getxattr(2) should return -1 and set
Basic tests for setxattr(2) and make sure setxattr(2) handles error
conditions correctly.
There are 7 test cases:
1. Any other flags being set except XATTR_CREATE and XATTR_REPLACE,
setxattr(2) should return -1 and set errno to EINVAL
2. With XATTR_REPLACE flag set but the attribute does not e
On 07/28/2011 12:10 AM, Cyril Hrubis wrote:
> Hi!
>> +#define RUSAGE_THREAD 1
>
> Shouldn't this be set in system headers?
>
> And even if not, it probably would be on newer systems so better way
> should be:
>
> #ifndef RUSAGE_THREAD
> #define RUSAGE_THREAD 1
> #endif
corrected.
>
>> +static
getrusage04 - accuracy of getrusage() with RUSAGE_THREAD
This program is used for testing the following upstream commit:
761b1d26df542fd5eb348837351e4d2f3bc7bffe.
getrusage() returns cpu resource usage with accuracy of 10ms
when RUSAGE_THREAD is specified to the argument who. Meanwhile,
accuracy
Hi Cyril:
Can you please commit this patch set?
Thanks a lot
-Wanlong Gao
> Hi Cyril, Garrett:
>
> I have resend the patch set of Tang with some format fixes.
>
> If you have any comment with this version, please reply me,
> and I'll fix and resend, otherwise, please commit them.
>
>
>
> T
On Dec 22, 2011, at 2:11 AM, Cyril Hrubis wrote:
> Hi!
testcase create NCPU=4 threads running with a high priority with the same
sched
policy and one with a low-priority. The low-priority thread should not run
until the other threads stop running, unless the machine has more t
Hi!
> >
> > I think using cat /proc/cpuinfo get core number is better.
>
>
> But /proc/cpuinfo may not be supported by other OS.
Yes, procfs is linux specific.
--
Cyril Hrubis
chru...@suse.cz
--
Write once. Port to m
Hi!
> >> testcase create NCPU=4 threads running with a high priority with the same
> >> sched
> >> policy and one with a low-priority. The low-priority thread should not run
> >> until the other threads stop running, unless the machine has more than NCPU
> >processors
> >> when cpu core >4,testcas
On 12/22/2011 04:49 PM, Caspar Zhang wrote:
>
> On powerpc platform, getrusage03_child returns the following error
> message:
>
> getrusage03_child1 TBROK : OPTION PARSING ERROR - parse_opts: ERROR -
> Only short options are allowed
>
> This is because child_options[] didn't end up with
On powerpc platform, getrusage03_child returns the following error
message:
getrusage03_child1 TBROK : OPTION PARSING ERROR - parse_opts: ERROR -
Only short options are allowed
This is because child_options[] didn't end up with a NULL array. This
patch fixes this problem.
Signed-off-by:
On 12/22/2011 04:31 PM, Caspar Zhang wrote:
>
> thp02 - detect mremap bug when THP is enabled.
>
> There was a bug in mremap THP support, sometimes crash happened
> due to the following reason according to developers:
>
> "alloc_new_pmd was forcing the allocation of a pte before calling
> move_h
thp02 - detect mremap bug when THP is enabled.
There was a bug in mremap THP support, sometimes crash happened
due to the following reason according to developers:
"alloc_new_pmd was forcing the allocation of a pte before calling
move_huge_page and that resulted in a VM_BUG_ON in move_huge_page
On 11/15/2011 09:14 PM, Zhouping Liu wrote:
> hi, Cyril & Garrett
> v5 -> v6: updated the sscanf() function in read_meminfo() in order to avoid
> buffer overflow.
>
> please feel free to comment the patch.
> BTW: you would miss the patch.
>
> There are several vm tunable files under /proc/sys/vm
On 11/15/2011 09:19 PM, Zhouping Liu wrote:
> hi, Cyril & Garrett
>
> please feel free to comment the patch.
>
> There are two tunables overcommit_memory and overcommit_ratio under
> /proc/sys/vm/, which can control memory overcommitment.
>
> The overcommit_memory contains a flag that enables me
34 matches
Mail list logo
