qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

ltp vs linux-user (was: Re: [Qemu-devel] [PATCH 2/2] check for utimensat


From: Rtp
Subject: ltp vs linux-user (was: Re: [Qemu-devel] [PATCH 2/2] check for utimensat() availability on configure)
Date: Fri, 12 Jun 2009 10:30:41 +0200
User-agent: Gnus/5.11 (Gnus v5.11) Emacs/22.2 (gnu/linux)

Riku Voipio <address@hidden> writes:

Hi,

>> > Generally we use tests from ltp testsuite since they have nice tests for 
>> > most
>> > syscalls. However, the testing is not quite systematic, an particularry 
>> > combinations
>> > of host/target glibc appear to create a mess..
>
>> Do you have a list of the syscalls/ltp testcases you're testing ? I've
>> made a list, only wants to compare :)
>
> The ltp tests realated syscalls being added/modified in a patch. As said, 
> there is
> no systematic "these tests must get through" list used at the moment. perhaps 
> we can
> use your list as starters?

Here is my list of syscalls tested. It's the same list as in
runtest/syscalls with the following tests disabled :

- clone*
- msgctl07
- profil101
- syslog*
- times03
- waitpid*

I've disabled them because they were blocking except for syslog
ones. The syslog ones were disabled because I didn't have any syslog
daemon in my chroot. I didn't look at the other tests but this can be
due to emulation bug or host bug. I need to run the same tests on the
host to be able to detect what's really due to qemu emulation.
(Please note that I was testing arm-linux-user on x86 host. I hope to
I'll find some time to do more tests).

Arnaud

Attachment: list
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]