sunhme lockup and udelay workaround

sunhme lockup and udelay workaround

[Daniel Drake]

Hi Dave,

Rewind a few years, I don't suppose you remember the sunhme lockup 
issue, of which one ugly workaround was inserting a small delay in the 
TX codepath?

http://dev.gentoo.org/~dsd/genpatches/trunk/2.6.19-pre/1705_sparc-U1-hme-lockup.patch

Was this ever resolved? This is the last patch in Gentoo's base set of 
kernel patches and I'm eager to remove it...

There is scattered discussion of this bug lying around, e.g.
http://lists.debian.org/debian-sparc/2003/02/msg00011.html

but it's not clear if a mainline resolution was ever reached. Do you 
know anything about the situation here?

Thanks!
Daniel

Re: sunhme lockup and udelay workaround

[David Miller]

From: Daniel Drake <dsd@gentoo.org>
Date: Sat, 18 Nov 2006 16:57:24 -0500

> Rewind a few years, I don't suppose you remember the sunhme lockup 
> issue, of which one ugly workaround was inserting a small delay in the 
> TX codepath?
> 
> http://dev.gentoo.org/~dsd/genpatches/trunk/2.6.19-pre/1705_sparc-U1-hme-lockup.patch
> 
> Was this ever resolved? This is the last patch in Gentoo's base set of 
> kernel patches and I'm eager to remove it...
> 
> There is scattered discussion of this bug lying around, e.g.
> http://lists.debian.org/debian-sparc/2003/02/msg00011.html
> 
> but it's not clear if a mainline resolution was ever reached. Do you 
> know anything about the situation here?

I've never seen this problem reappear since gcc-3.x or so.
I am pretty sure you can kill that patch safely.