Re: komd issues - was [Re: [openMosix-general] openmosix 2.4.26 in 2.4.31?]

On Thu, Sep 21, 2006 at 09:40:08AM -0600, Matt Dew wrote:
> Setting LD_ASSUME_KERNEL=2.4.1 fixes the system call segfaults.
> WOOHOO!! Now we know where to look to fix the problem.

TLS screwup, through fs/gs problem. more than likely.

> Unfortunately, I don't really understand LD_ASSUME_KERNEL in great
> detail.  

only details useful, is that is turn OFF the use of VDSO/TLS .. all
kind of fancy stuff..

00000000 <__getpid>:
0:	b8 14 00 00 00       	mov    $0x14,%eax
5:	65 ff 15 10 00 00 00 	call   *%gs:0x10
c:	c3                   	ret    

if %gs is not correct here, it's going to segfault.

> Are we going to have to redo the sys_call_table?  Have each entry call a
> new om_sys_call_table that calls the appropriate function?

The syscall_table is absolutely constant. only entry point changes.

Note that i386 sysenter type of syscall (intel syscall on p4)
have never been tested. I changed the entry point code there, but never
had any power-sucker-processor to test it.

-- 
Vincent Hanquez

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV