[Top] [All Lists]

Re: MIPS 2.1.89 now in CVS

To: Thomas Bogendoerfer <>,,,
Subject: Re: MIPS 2.1.89 now in CVS
Date: Sun, 22 Mar 1998 20:21:40 +0100
In-reply-to: <>; from Thomas Bogendoerfer on Sun, Mar 22, 1998 at 05:21:50PM +0100
References: <>
> irgendwie mag die zugegeben durch unser Mailsystem verunstaltete
> (aber IHMO gueltige) From Adresse nicht. Da Anwort/Frage primaer dich
> betreffen, schick ich sie nur an dich.

I don't have the holy RFC at hand but if I remember it right unbalanced
violate the spec.  Anyway, if you want to very if your mail is RFC down
to the letter, try mailing Ralph Babel.  If you mail bounces, it isn't  ;-)

> On Tue, Mar 17, 1998 at 11:48:43PM +0100, wrote:
> > I've upgraded the MIPS port to 2.1.89.  Aside of the still unresolved
> > problem in exit_mm() which I've reported some hours ago is seems to be
> > working on Indys.  Haven't yet tested the rest of my hardware collection.
> Is the exit_mm() problem still unresolved ?

I solved it.  There is a couple of superfluous fput calls in fs/binfmt_elf.c
Will commit the fix later today along with other bug fixes and performance
improvments.  Still haven't verified if this is a bug in our private
version or in Linus' version as well.

The bad new is that I still haven't been able to track down the case of
memory corruption that is plaguing the Indy.

> > I'm particularly interested in reports about the NCR53C9x driver on
> > Olivetti M700 / Magnum 4000.
> works so far. But I can't say much, because fsck bombs out due to the
> exit_mm() problem. But detection of the scsi devices and mounting the root
> filesystem works.

Ok.  The Indy driver also survived the switch to the new SCSI midlevel
code, so everything went alot smother than initially reading the SCSI
code made me believe.


<Prev in Thread] Current Thread [Next in Thread>