This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug testsuite/10007] Some SystemTap tests need updates to handle SYSCALL_WRAPPERS


------- Additional Comments From ananth at in dot ibm dot com  2009-03-26 14:13 -------
Subject: Re:  Some SystemTap tests need updates to
	handle SYSCALL_WRAPPERS

On Thu, Mar 26, 2009 at 01:56:26PM -0000, dsmith at redhat dot com wrote:
 
> I took a short look at this patch.  Besides the testsuite changes, it looks like
> there is also an actual s390 syscall tapset change in there.

Thanks Dave! That small s390 syscall tapset change was to accomodate for
the missing sys32_mmap and old_mmap syscalls, which were failing to
resolve correctly. I don't have access to a s390 box right now to verify
exactly what their substitutes are. So, I envisage there may be a few
more changes once I am able to correctly fix the s390 specific issue.

> I'm a bit confused by the changes to systemtap.base/stmtvars.exp and
> systemtap.base/system_func.stp.  Instead of changing
> 'kernel.function("sys_open")' to 'syscall.open', you changed it to
> 'kernel.function("do_sys_open")'.  Why was this done?

Again, syscall wrappers issue. sys_open() will only exist as an inline
and SyS_open is what we need there. However, for that particular case, I
found it prudent to just change the probed function to one level down
(do_sys_open is called from sys_open irrespective of syscall wrappers).

> Other than those points, I guess I'm a bit worried that syscall tapset bugs are
> going to cause lots of (somewhat unrelated) testsuite failures.  But, I don't
> see any easier solution to this problem.

Right. I wouldn't be surprised if the more mainstream 64bit
architectures also gain this feature in the future.


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=10007

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]