On 08/17/2015 04:17 PM, Stanislav Kholmanskikh wrote:
>
>
> On 08/17/2015 03:49 PM, Jan Stancek wrote:
>>
>>
>>
>>
>> - Original Message -
>>> From: "Stanislav Kholmanskikh"
>>> To: ltp-list@lists.sourceforge.net
>
On 08/17/2015 03:49 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: ltp-list@lists.sourceforge.net
>> Cc: "vasily isaenko"
>> Sent: Monday, 17 August, 2015 2:13:17 PM
>>
Makefiles in kernel/containers explicitly enumerate all
required libraries in LDLIBS. This avoids linking the
test cases with additional libraries, whereas all other LTP
test cases can be linked this way.
Fixed this by using $(LDLIBS) instead of '-lltp'.
Signed-off-by: Stanislav Kh
uction we
need ('mov ...').
The trick was proposed by Jose E. Marchesi .
Signed-off-by: Stanislav Kholmanskikh
---
include/lapi/rt_sigaction.h | 20
1 files changed, 12 insertions(+), 8 deletions(-)
diff --git a/include/lapi/rt_sigaction.h b/include/lapi/rt_sigactio
On 06/11/2015 02:57 PM, Cyril Hrubis wrote:
> Hi!
> Acked.
>
Thank you. Pushed.
--
___
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/li
On 06/10/2015 03:40 PM, Maninder Singh wrote:
> As in commit 9fa8ad0 Change tst_resm() followed by tst_exit() to tst_brkm()
>
> Signed-off-by: Maninder Singh
> Signed-off-by: Akhilesh Kumar
> ---
> testcases/kernel/syscalls/flock/flock01.c | 12 ++--
> testcases/kernel/syscalls/flo
0
close(3) = 0
open("/dev/loop0", O_RDWR) = 3
open("file.img", O_RDWR) = 4
ioctl(3, LOOP_SET_FD, 0x4) = -1 EBUSY (Device or resource busy)
So let's wait until the kernel frees the loop device, i.e.
when ioctl(LOOP_CLR_FD)
On 06/10/2015 01:30 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Cyril Hrubis"
>> To: "Stanislav Kholmanskikh"
>> Cc: ltp-list@lists.sourceforge.net, "vasily isaenko"
>> , jstan...@redhat.com
On 06/10/2015 01:08 PM, Cyril Hrubis wrote:
> Hi!
>> Signed-off-by: Stanislav Kholmanskikh
>> ---
>> lib/tst_device.c | 17 ++---
>> 1 files changed, 6 insertions(+), 11 deletions(-)
>>
>> diff --git a/lib/tst_device.c b/lib/tst_de
Hi.
On 06/10/2015 10:58 AM, Maninder Singh wrote:
> As in commit 9fa8ad0 Change tst_resm() followed by tst_exit() to tst_brkm()
>
> Signed-off-by: Maninder Singh
> Signed-off-by: Akhilesh Kumar
> ---
> testcases/kernel/syscalls/flock/flock01.c | 12 ++--
> 1 file changed, 2 insertio
On 06/10/2015 09:56 AM, Pratik Prajapati wrote:
> Hi,
Hi.
>
> I want to add a test case for my local use. From
> https://github.com/linux-test-project/ltp/wiki/Test-Writing-Guidelines I
> get to know how to write a test case but didb't get how to add it.
> Please help.
For example, you could c
On 06/09/2015 03:50 PM, Cyril Hrubis wrote:
>
>
> I think that we should do:
>
> detach_device(..)
> {
> while EBUSY:
> ioctl(fd, LOOP_CLR_FD);
>
> while not error:
> ioclt(LOOP_GET_STATUS);
> usleep();
> }
>
> Because if we ignore the return
On 06/10/2015 11:11 AM, Stanislav Kholmanskikh wrote:
On systems employing the "lazy loop device removal",
kernel commit commit a1ecac3b0656a68259927c234e505804d33a7b83
("loop: Make explicit loop device destruction lazy"),
ioctl(LOOP_CLR_FD) may return 0 when the loop d
op device, i.e.
when ioctl(LOOP_CLR_FD) starts failing with ENXIO.
Signed-off-by: Stanislav Kholmanskikh
---
lib/tst_device.c | 17 ++---
1 files changed, 6 insertions(+), 11 deletions(-)
diff --git a/lib/tst_device.c b/lib/tst_device.c
index 3bded53..9a92fc8 100644
--- a/lib/tst_
On 06/09/2015 03:19 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: "Cyril Hrubis" , "Jan Stancek"
>> Cc: ltp-list@lists.sourceforge.net, "vasily isaenko"
>>
On 06/09/2015 03:03 PM, Cyril Hrubis wrote:
> Hi!
>>> We should probably wrap the su into a script that makes sure that the
>>> return value is compatible with LTP. Because it may set bogus test
>>> return status if su reported anything else than 0 or 1.
>>
>> Ehm, but -c "$0" will execute the sa
On 06/09/2015 01:54 PM, Cyril Hrubis wrote:
> Hi!
>> Now we verify the exit code of the second test case. We did not
>> do this previously, and it was a bug.
>>
>> Signed-off-by: Stanislav Kholmanskikh
>> ---
>> testcases/kernel/fs/acls/acl_test01 |3
On 06/09/2015 01:10 PM, Cyril Hrubis wrote:
> Hi!
>> If the test is executed on NFS, it may print:
>>
>> rm: cannot remove `/mnt/ltp-xeJYpESVKz/acltest01.fCogoo4gn0': Directory not
>> empty
>>
>> It seems the following happen:
>>
>> 1. After 'unmount -d' 'acltest01.fCogoo4gn0/blkext3' is still i
Hi.
On 06/09/2015 01:00 PM, Cyril Hrubis wrote:
> Hi!
>> -
>> -tst_require_root
>
> Why do we remove this? The test calls adduser and we need root for that
> or don't we?
This is because we call 'tst_require_root' twice. The first time is from
"FUNCTION MAIN:" and the second time when we exe
n0' directory fails, because it
still contains this '.nfs*' file.
Intorucing a wait cycle should fix this problem.
Signed-off-by: Stanislav Kholmanskikh
---
It's a follow up to:
http://sourceforge.net/p/ltp/mailman/message/32987268/
Indeed, 'sync' was wro
Converted the test case to use tst_tmpdir instead of local juggling
with $TMP.
When we execute LTP, $PATH includes $LTPROOT/testcases/bin, so there
is no reason to set $TCbin explicitly. Therefore, removed $TCbin.
Signed-off-by: Stanislav Kholmanskikh
---
runtest/admin_tools
Signed-off-by: Stanislav Kholmanskikh
---
doc/test-writing-guidelines.txt | 17 +
tools/apicmds/.gitignore|1 +
tools/apicmds/Makefile |2 +-
tools/apicmds/ltpapicmd.c | 17 +
4 files changed, 36 insertions(+), 1 deletions
Now we verify the exit code of the second test case. We did not
do this previously, and it was a bug.
Signed-off-by: Stanislav Kholmanskikh
---
testcases/kernel/fs/acls/acl_test01 |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/testcases/kernel/fs/acls/acl_test01
b
The output from 'grep' gives no value, so let's hide it.
'losetup -d' is also not needed, because the loop device is
detached by 'umount -d'. Therefore, removed 'losetup -d' to hide
warnings like:
loop: can't delete device /dev/loop1: N
On 05/26/2015 01:21 PM, Stanislav Kholmanskikh wrote:
> As of now, test cases utilizing tst_checkpoint framework fail on NFS
> with similar errors:
>
> creat07 0 TWARN : tst_tmpdir.c:206: tst_rmdir:
> rmobj(/tmpdir/ltp-0IVQPP0NK6/creigi0kV) failed:
> remove(/tmp
keep the futex page mapped while tst_rmdir() is running.
Let's unmap it before the actual delete procedure.
Signed-off-by: Stanislav Kholmanskikh
---
Changes since V1:
* renamed 'futex_t *futexes' to 'futex_t *tst_futexes'
lib/tst_checkpoint.c | 12 +++-
li
keep the futex page mapped while tst_rmdir() is running.
Let's unmap it before the actual delete procedure.
Signed-off-by: Stanislav Kholmanskikh
---
lib/tst_checkpoint.c |2 +-
lib/tst_tmpdir.c | 11 +++
2 files changed, 12 insertions(+), 1 deletions(-)
diff --git
Signed-off-by: Stanislav Kholmanskikh
---
include/lapi/futex.h| 24
lib/tst_checkpoint.c|3 +--
testcases/kernel/syscalls/futex/futextest.h |2 +-
3 files changed, 26 insertions(+), 3 deletions(-)
create mode
On 05/25/2015 11:51 AM, Cyril Hrubis wrote:
> Hi!
>> As of now, if we run test cases which use the tst_checkpoint framework on
>> NFS,
>> they will fail with errors like:
>>
>> creat07 0 TWARN : tst_tmpdir.c:206: tst_rmdir:
>> rmobj(/tmpdir/ltp-0IVQPP0NK6/creigi0kV) failed:
>> remove(/t
empty
This happens because we keep the futex page mapped while tst_rmdir() is running.
So on we need to unmap the page before tst_rmdir().
What do you think about the proposed solution?
Thanks.
Signed-off-by: Stanislav Kholmanskikh
---
include/tst_checkpoint.h |9
ere to use number of present CPUs, rather than all
>
> Changes in v3:
> + use present_mask variable in get_present_cpus
> + no special treatment for cpu0 in cpuhotplug03
> + fix cpuhotplug03 to spawn (number of present cpus)*2 processes
> + this time whole series has correct version in subje
On 04/29/2015 05:41 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: "Jan Stancek" , ltp-list@lists.sourceforge.net
>> Sent: Wednesday, 29 April, 2015 3:09:06 PM
>> Subject: Re:
r sparc64 kernel. To be
honest, there is no cpu hotplug (yet), so all this test cases return
TCONF ;)
>
> Regards,
> Jan
>
>
> Jan Stancek (1):
>cpu_hotplug: use hotplug/present cpus functions
>
> Stanislav Kholmanskikh (3):
>cpuhotplug: use cpu states in cleanu
On 04/29/2015 02:18 PM, Jan Stancek wrote:
> Signed-off-by: Jan Stancek
> ---
> testcases/kernel/hotplug/cpu_hotplug/functional/cpuhotplug01.sh | 6 +++---
> testcases/kernel/hotplug/cpu_hotplug/functional/cpuhotplug02.sh | 2 +-
> testcases/kernel/hotplug/cpu_hotplug/functional/cpuhotplug03.s
On 04/29/2015 02:17 PM, Jan Stancek wrote:
> From: Stanislav Kholmanskikh
>
> Signed-off-by: Stanislav Kholmanskikh
> Signed-off-by: Jan Stancek
> ---
> .../cpu_hotplug/include/cpuhotplug_hotplug.sh | 73
> +-
> 1 file changed, 70 in
Hi.
On 04/20/2015 01:28 PM, Matus Marhefka wrote:
> Hi,
>
> please consider pushing this reviewed patch.
>
> Thanks,
> Matus Marhefka
>
>
> - Original Message -
> From: "Jan Stancek"
> To: "Matus Marhefka"
> Cc: ltp-list@lists.sourceforge.net
> Sent: Thursday, February 26, 2015 2:22:24 P
On 04/15/2015 04:13 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: ltp-list@lists.sourceforge.net
>> Cc: "vasily isaenko"
>> Sent: Wednesday, 15 April, 2015 11:27:09 AM
>>
In order to put output messages in order and remove "duplicates",
we need a fflush() before forking and tst_fork() does it.
Signed-off-by: Stanislav Kholmanskikh
---
testcases/kernel/mem/tunable/max_map_count.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a
On 04/14/2015 12:19 PM, Cyril Hrubis wrote:
> Hi!
We have several issues in our sparc64 environment. Need a day or two to
figure out whether they are test case issues or problems with our
environment.
>>>
>>> Ok.
>>>
>>
>> Finally, there are only two test case issues:
>>* cpuho
On 04/07/2015 06:27 PM, Cyril Hrubis wrote:
> Hi!
It's about the time we start to prepare for next release. I will
(hopefully) start runing latest git code and look for unexpected
failures during this week. Everybody please try to run the latest
git code and report any problem
Signed-off-by: Stanislav Kholmanskikh
---
include/lapi/semun.h | 35
.../kernel/ipc/ipc_stress/semaphore_test_01.c | 13 +---
.../kernel/ipc/ipc_stress/semaphore_test_02.c |7 +---
.../kernel/ipc/ipc_stress
It's required by the semctl() API plus without that the test
case may fail on 64-bit Big Endian systems (like sparc64).
Signed-off-by: Stanislav Kholmanskikh
---
testcases/kernel/containers/sysvipc/sem_comm.c |9 +++--
1 files changed, 7 insertions(+), 2 deletions(-)
diff --
Signed-off-by: Stanislav Kholmanskikh
---
.../hotplug/cpu_hotplug/functional/cpuhotplug04.sh | 56 +++
1 files changed, 32 insertions(+), 24 deletions(-)
diff --git a/testcases/kernel/hotplug/cpu_hotplug/functional/cpuhotplug04.sh
b/testcases/kernel/hotplug/cpu_hotplug/functional/
Hi!
The current version of cpuhotplug04.sh fails with our sparc64 kernel:
Name: cpuhotplug04
Date: Mon Apr 6 14:30:45 EDT 2015
Desc: Does it prevent us from offlining the last CPU?
cat: /sys/devices/system/cpu/cpu3/online: No such file or directory
/opt/ltp/testcases/bin/cpuhotplug_hotplu
ff-by: Stanislav Kholmanskikh
---
.../hotplug/cpu_hotplug/functional/cpuhotplug01.sh | 11 ---
.../hotplug/cpu_hotplug/functional/cpuhotplug03.sh | 12 +---
.../hotplug/cpu_hotplug/functional/cpuhotplug04.sh | 18 +-
.../cpu_hotplug/include/cpuhotplug_hotpl
Signed-off-by: Stanislav Kholmanskikh
---
.../cpu_hotplug/include/cpuhotplug_hotplug.sh | 22
1 files changed, 22 insertions(+), 0 deletions(-)
diff --git a/testcases/kernel/hotplug/cpu_hotplug/include/cpuhotplug_hotplug.sh
b/testcases/kernel/hotplug/cpu_hotplug
On 04/02/2015 01:09 PM, Cyril Hrubis wrote:
> Hi!
>> It's about the time we start to prepare for next release. I will
>> (hopefully) start runing latest git code and look for unexpected
>> failures during this week. Everybody please try to run the latest
>> git code and report any problems.
>
> I'v
Hi!
On 02/16/2015 12:22 PM, Wei,Jiangang wrote:
> These mistakes cann't lead to trouble for reading,
> but the correct spelling is more better for beginners
> to understand it.
>
> Signed-off-by: Wei,Jiangang
Thank you, pushed.
--
Hi!
On 02/12/2015 07:59 PM, Cyril Hrubis wrote:
> Hi!
> Code prototype follows. It uses mutexes and can be used with up to
> page_size/sizeof(uint32) pairs. I've checked that it compiles fine and
> works back to kernel 2.6.11 compiled in year 2005, which should be more
> than enough.
>
> It uses a
On 02/03/2015 07:12 PM, Cyril Hrubis wrote:
> Hi!
>> Signed-off-by: Stanislav Kholmanskikh
>> ---
>> Changes since V1:
>> * Applied all comments from V1
>> * Added one more const to cmd[] in Example
>
> Acked.
>
Thank you for review.
tst_run_cm
Signed-off-by: Stanislav Kholmanskikh
---
Changes since V1:
* Applied all comments from V1
* Added one more const to cmd[] in Example
doc/test-writing-guidelines.txt | 40 +++
1 files changed, 40 insertions(+), 0 deletions(-)
diff --git a/doc/test
On 02/03/2015 05:35 PM, Cyril Hrubis wrote:
> Hi!
>> +2.2.20 Executing a file
>> +^^^
>
> Maybe "Running executables" would be a bit better, as it is the title
> looks strange to me.
I stole it from 'man 3 execvp' :) No problem:)
>
>> +[source,c]
>> +
It may happen, that the first execution of 'rmmod module_name'
may fail for some reason (like the module is in use for some short time).
Changed tst_module_unload() to execute 'rmmod module_name' multiple
times with the total number limited by a timeout.
Signed-off-by: St
On 01/26/2015 04:11 PM, Cyril Hrubis wrote:
>> @@ -100,6 +153,21 @@ void tst_module_load(void (cleanup_fn)(void),
>>
>> void tst_module_unload(void (cleanup_fn)(void), const char *mod_name)
>> {
>> +int i;
>> +int loaded;
>> +
>> const char *const argv[] = { "rmmod", mod_name, NU
Signed-off-by: Stanislav Kholmanskikh
---
Sorry for late response.
doc/test-writing-guidelines.txt | 27 +++
1 files changed, 27 insertions(+), 0 deletions(-)
diff --git a/doc/test-writing-guidelines.txt b/doc/test-writing-guidelines.txt
index 4d70a1b..a0004d8 100644
Added an option to pass the program exit code to the caller.
This may be needed if we want to run a program with tst_run_cmd(),
but handle failures manually (i.e. not call cleanup).
Signed-off-by: Stanislav Kholmanskikh
---
include/test.h | 16 ++---
lib
It may happen, that the first execution of 'rmmod module_name'
may fail for some reason (like the module is in use for some short time).
Changed tst_module_unload() to execute 'rmmod module_name' multiple
times with the total number limited by a timeout.
Signed-off-by: St
Hi, and thank you for review.
On 01/19/2015 05:52 PM, Cyril Hrubis wrote:
> Hi!
>> There is one more issue with tbio. :(
>>
>> Sometimes it may fail with:
>>
>> ltp_tbio0 TINFO : Device opened successfully
>> ltp_tbio1 TPASS : success on LTP_TBIO_ALLOC test
>> ltp_tbio2 TPASS :
n_cmd_fds are
acceptable.
Thanks!
Signed-off-by: Stanislav Kholmanskikh
---
include/test.h |8 ++-
include/tst_module.h | 11
lib/tst_mkfs.c |2 +-
lib/tst_module.c
* Wait until udev creates /dev/tbio. If udev can not create
the node file in a defined timeout, create it manually.
* Clean /dev/tbio after test execution, if /dev/tbio was created
manually.
Signed-off-by: Stanislav Kholmanskikh
---
Changes since V2:
* Check for the file ten times a second
* Wait until udev creates /dev/tbio. If udev can not create
the node file in a defined timeout, create it manually.
* Clean /dev/tbio after test execution, if /dev/tbio was created
manually.
Signed-off-by: Stanislav Kholmanskikh
---
Changes since V1:
* implemented a more active waiting
Hi!
On 01/06/2015 01:18 PM, Cyril Hrubis wrote:
> Hi!
>> +/*
>> + * Wait until udev creates the device node.
>> + * If the node is not created, we will create it manually,
>> + * see below.
>> + */
>> +sleep(5);
>
> As usually I do not like adding long sleep() into testcase
* Introduce a sleep() to wait until udev creates /dev/tbio
* Fix typos
* Clean /dev/tbio after test execution, if /dev/tbio was created
manually.
Signed-off-by: Stanislav Kholmanskikh
---
.../kernel/device-drivers/tbio/tbio_user/tbio.c| 20 +++-
1 files changed, 15
inter->cmd containing a null-terminated
string,
and inter->cmd_len containing the exact number of bytes to store this string.
Signed-off-by: Stanislav Kholmanskikh
---
.../device-drivers/tbio/tbio_kernel/ltp_tbio.c | 11 ---
.../kernel/device-drivers/tbio/tbio_kernel/tbio.h |2
In some of our environments (x86, x86_64, sparc64) tbio fails,
and this series fixes the failures.
I tried to introduce minimal changes, because, to be honest, I'm not
100% sure that I understand the block layer very well. So I would be
more than happy, if anybody will have a chance to review/test
tbio_open().
Signed-off-by: Stanislav Kholmanskikh
---
.../device-drivers/tbio/tbio_kernel/ltp_tbio.c |8 +++-
1 files changed, 3 insertions(+), 5 deletions(-)
diff --git a/testcases/kernel/device-drivers/tbio/tbio_kernel/ltp_tbio.c
b/testcases/kernel/device-drivers/tbio/tbio_kernel/
Hi, George.
I've just found that I'm also motivated to have this bug fixed :)
Are you planning to send an updated version including Jan's comments?
Thanks.
PS: I also tested it with access06 test case:
without the patch:
[root@kholmanskikh access]# LTP_DEV_FS_TYPE=btrfs ./access06
access06
On 11/28/2014 11:33 AM, Han Pingtian wrote:
> On Mon, Nov 24, 2014 at 05:53:32PM +0300, Stanislav Kholmanskikh wrote:
>> Hi.
>>
>> In general, it looks good.
>>
>> I have only one comment, that it's possible that the uid/gid values we pass
>> to t
d, Nov 19, 2014 at 11:15:54AM +0300, Stanislav Kholmanskikh wrote:
>>> Looks like those cases need to be fixed for the 16-bit compat, right?
>>>
>>> chown0[1-5] fchown0[1-5] lchown03 setresgid0[1-4] setresuid0[1-3,5]
>>
>> Looks so.
>>
>> Sorry for
On 11/13/2014 10:57 AM, Han Pingtian wrote:
> On Mon, Nov 10, 2014 at 12:07:20PM +0300, Stanislav Kholmanskikh wrote:
>> Hi!
>>
>> On 10/31/2014 04:39 PM, Han Pingtian wrote:
>>> Hey there,
>>>
>>> I found that looks like all chown test cases will
Hi!
On 10/31/2014 04:39 PM, Han Pingtian wrote:
> Hey there,
>
> I found that looks like all chown test cases will be compiled to 16 bit
> compat test case, but they just call chown() something like
>
> TEST(chown(fname, uid, gid));
>
> which won't test if there is a 16 bit chown() syscall. I
explicit loop device destruction lazy
To cope with this test case failure I introduced a 'sync'.
And 'losetup -d' seems to be unnecessary, because of 'umount -d' before it
Signed-off-by: Stanislav Kholmanskikh
---
testcases/kernel/fs/acls/acl_test01 |2 +-
1 files chan
Hi!
Patch 3 was pushed.
Thank you.
On 09/28/2014 06:53 AM, Xiong Zhou wrote:
> Hi!
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: "Xiong Zhou"
>> Cc: ltp-list@lists.sourceforge.net
>> Sent: Friday, September 26,
Hi!
On 09/26/2014 07:23 AM, Xiong Zhou wrote:
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: "Xiong Zhou"
>> Cc: ltp-list@lists.sourceforge.net
>> Sent: Thursday, September 25, 2014 8:07:26 PM
>> Subject: Re
Hi!
On 09/25/2014 11:17 AM, Xiong Zhou wrote:
>
> And block3 description message fix from "mandatory locking"
> to "negative whence".
>
> Signed-off-by: Xiong Zhou
> ---
Patch 1 and patch 2 of the series are pushed.
Btw, patch 2 didn't apply to index correctly with 'git am', I had to
manually a
Hi!
On 09/19/2014 10:46 AM, Xiong Zhou wrote:
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: "Xiong Zhou" , ltp-list@lists.sourceforge.net
>> Sent: Thursday, September 11, 2014 10:04:32 PM
>> Subject: Re:
Hi!
On 09/19/2014 10:49 AM, Xiong Zhou wrote:
>
> Tests splice01 tee01 and utime01/02/03 pass on NFS since 2.6.32 at least.
Could you add a kernel version check to the patch, please?
I'm sure there are distributions which still use older kernels.
For example, with 2.6.18-371.3.1.0.1.el5 and:
sp
On 09/19/2014 05:20 PM, chru...@suse.cz wrote:
> Hi!
>> Could anybody help to review this change, please?
>
> Sorry, I've read the email but forgot to reply the last time.
>
> Looks good, acked.
>
Thank you. Pushed.
--
M
Hi!
Could anybody help to review this change, please?
Thanks.
On 21.08.2014 12:44, Stanislav Kholmanskikh wrote:
> I think it's not needed to use set_block_device(),
> create_block_device() from runltp, because their logic
> is in tst_acquire_device()/tst_release_device() now.
>
Hi!
On 09/11/2014 02:32 PM, Xiong Zhou wrote:
> NFS support splice(2) tee(2) and utime(2) now.
I think that 'now' is a bit relative. I.e. for me 'now'
can be 2.6.39, but for another guy - 3.16.2 or whatever.
My proposal:
* if this support came many-many years ago, we just mark the fact in
the
tst_resm(TINFO, "Exit block 2");
> + } else
> + tst_resm(TINFO, "Skip block 2 on NFS");
It's more a matter of taste, but recently in commit:
commit a2d860c854511cccd682ce244636cbb5f4b75d7c
Author: Xiong Zhou
Date: Tue
Hi!
On 09/11/2014 02:30 PM, Xiong Zhou wrote:
> And block3 description message fix from "mandatory locking"
> to "negative whence".
>
> Signed-off-by: Xiong Zhou
> ---
> testcases/kernel/syscalls/fcntl/fcntl14.c | 36
> +++
> 1 file changed, 13 insertions(+), 23 de
On 09/08/2014 01:12 PM, Stanislav Kholmanskikh wrote:
> Hi.
>
> Could anybody look at the series?
>
> If there are no objections, I'm going to push it tomorrow.
Pushed.
>
> Thanks.
>
> On 08/25/2014 12:26 PM, Stanislav Kholmanskikh wrote:
>>
>>
Hi.
Could anybody look at the series?
If there are no objections, I'm going to push it tomorrow.
Thanks.
On 08/25/2014 12:26 PM, Stanislav Kholmanskikh wrote:
>
>
> On 08/22/2014 11:05 AM, Xiong Zhou wrote:
>>
>> According to nfs(5), NLM supports advisory file lock
no need to resend the
patch/series because it can be fixed at commit time. :)
Thanks.
Reviewed-by: Stanislav Kholmanskikh
> testcases/kernel/syscalls/fcntl/fcntl16.c | 27 +--
> 1 file changed, 21 insertions(+), 6 deletions(-)
>
> diff --git a/testcases/k
> or not aligned on a page boundary).", which suggests that both
>> ENOMEM and EINVAL should be considered.
>>
>> Signed-off-by: Jan Stancek
>
> Acked-by: Wanlong Gao
Modern sparc64 kernels return EINVAL as well.
Acked-by: Stanislav Kholmanskikh
>
>> ---
>
Hi!
On 08/21/2014 08:12 AM, Xiong Zhou wrote:
>
> According to nfs(5), NLM supports advisory file locks only. So skip
> fcntl16 test if NFS.
>
I've just took a fresh look at fcntl16.c and found that block1 of this
test case may be executed on NFS. And, IMHO, this block1 test case is
not 100% i
On 08/21/2014 08:17 AM, Xiong Zhou wrote:
>
> According to description of NFS and directIO in open(2), especially
> "The Linux NFS client places no alignment restrictions on
> O_DIRECT I/O", ignore "odd count"/"non-aligned" read-write FAILs
> in diotest4.
>
> Signed-off-by: Xiong Zhou
> ---
Loo
On 08/21/2014 08:14 AM, Xiong Zhou wrote:
>
> Split dup code cleanup and function change from v2.
I suppose this message is directed to the LTP mailing list subscribers
and should not be in `git log` after the patch is applied. Therefore put
such messages after '---'.
One small comment below.
I think it's not needed to use set_block_device(),
create_block_device() from runltp, because their logic
is in tst_acquire_device()/tst_release_device() now.
Signed-off-by: Stanislav Kholmanskikh
---
testscripts/ltpstress.sh | 13 +++--
1 files changed, 11 insertions(+), 2 dele
Hi!
Sorry for delay.
On 08/12/2014 08:37 AM, Xiong Zhou wrote:
>
> According to description of NFS and directIO in open(2), especially
> "The Linux NFS client places no alignment restrictions on
> O_DIRECT I/O", ignore "odd count"/"non-aligned" read-write FAILs
> in diotest4. Some dup code clean
On 08/14/2014 03:50 PM, Jan Stancek wrote:
>
>
>
>
> - Original Message -
>> From: "Stanislav Kholmanskikh"
>> To: ltp-list@lists.sourceforge.net
>> Cc: "vasily isaenko"
>> Sent: Thursday, 14 August, 2014 10:30:51 AM
>>
1. Changed '__sparc64__' to 'defined(__sparc__) && defined(__arch64__)'
2. Removed non-implemented syscall numbers from sparc64.in (in accordance
to linux-stable/arch/sparc/include/uapi/asm/unistd.h)
2. Re-executed regen.sh
Signed-off-by: Stanislav Kholmanskikh
On 08/12/2014 04:24 PM, chru...@suse.cz wrote:
> Hi!
>> On some platforms LTP fails to compile, because (struct sigaction)
>> doesn't have .sa_restorer member. For example - ia64.
>>
>> To overcome such failures we employ HAVE_SA_RESTORER define.
>>
&g
On some platforms LTP fails to compile, because (struct sigaction)
doesn't have .sa_restorer member. For example - ia64.
To overcome such failures we employ HAVE_SA_RESTORER define.
Signed-off-by: Stanislav Kholmanskikh
---
Verified on SPARC, x86_64, i386.
include/lapi/rt_sigaction.h |
On 08/12/2014 01:04 PM, chru...@suse.cz wrote:
> Hi!
>>> I've started to prepare for release and tried to build LTP on as much
>>> arch/distro combination I could, for ia64 the rt_sigaction fails with:
>>>
>>> In file included from rt_sigaction01.c:42:
>>> [ 349s] ../../../../include/lapi/rt_sig
On 08/11/2014 07:04 PM, chru...@suse.cz wrote:
> Hi!
Hi!
> I've started to prepare for release and tried to build LTP on as much
> arch/distro combination I could, for ia64 the rt_sigaction fails with:
>
> In file included from rt_sigaction01.c:42:
> [ 349s] ../../../../include/lapi/rt_sigact
Hi!
On 08/11/2014 10:57 AM, Xiong Zhou wrote:
> According to description of NFS and directIO in open(2), especially
> "The Linux NFS client places no alignment restrictions on
> O_DIRECT I/O", ignore some FAILs in diotest4.
>
> According to nfs(5), NLM supports advisory file locks only. So skip
>
On 07/29/2014 02:11 PM, chru...@suse.cz wrote:
> Hi!
>> Sorry, I overlooked that thread "[PATCH v2 1/2] lib/tst_res.c: introduce
>> tst_record_childstatus()" already addresses these issues.
>
> And it took longer than it should because it require changes in the test
> library.
>
> I've just commi
On 07/25/2014 12:03 PM, Stanislav Kholmanskikh wrote:
>
>
> On 07/19/2014 01:35 AM, Joseph Beckenbach wrote:
>> Hi, all!
>
> Hi!
>
>>
>> Feel free to ignore this -- it's an experience report of a cluster of runs
>> I've just done agai
1 - 100 of 467 matches
Mail list logo
