Re: [linux-sh:01865] [linuxsh-dev] 2.4.7 unaligned user space access problems

[Stuart M. <Stu...@st...>]

Folks

I've not had a chance to test the new 2.4.7, but one failure we found a
few weeks ago may be relevant.

Just before the major revamp of the cache handling code started, an
optimisation was put into __flush_tlb_phys() which stoped things working
on our hardware when under heavy load.

This piece of code loops over TLB entries looking for ones to be flushed,
and this has to be done from P2. The optimisation was to only perform
the write to the memory mapped TLB registers from P2, and the rest of the
loop from P1 (previously the whole function was run from P1). Even though
this code followed all the guidelines in the SH7750 and SH7751 manuals, it
still appeared to cause cache corruption when the optimisation was in
place (we were seeing 'code' in the data cache, but I can't be any more
specific), and didn't using the previous version.

We never pinned this down any further, and its been on the TODO list for
a while now (although there is nothing to look at now!). However it
sounds like some of the optimisations which have been done recently may
be similar, so it might be worth a look.

Just an idea

Stuart



On Jul 24,  9:07am, du...@se... wrote:
> Subject: RE: [linux-sh:01865] [linuxsh-dev] 2.4.7 unaligned user space acc
> I've tried the new patch, but I'm still getting unaligned access faults.
> The faults seem to be generated in somewhat random locations within the
> bootup.  I don't know if 'stale' cache information is causing this, but it
> would explain the randomness of the errors.  I'll back up to 2.4.6 and look
> for these faults.
> 
> Dustin.
> 
> -----Original Message-----
> From: NIIBE Yutaka [mailto:gn...@m1...]
> Sent: Tuesday, July 24, 2001 3:59 AM
> To: linux-sh; linuxsh-dev
> Subject: [linux-sh:01865] [linuxsh-dev] 2.4.7 unaligned user space
> access problems
> 
> 
> I've tested some kernels.  My conclusion is: either CqREEK is broken
> or kernel is broken.
> 
> Folks, please test the current CVS version and if you have time test
> stock 2.4.6 kernel.  Then, please let me know it works well or not.
> 
> 
> CVS current (2.4.7+change):
>     CqREEK	SolutionEngine
>     Bad		OK
> 
> vanilla 2.4.7+atomic+ksoftirq:
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.6
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.6-pre6
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.6-pre4
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.6-pre2+fix(bitops,semaphore,irq.c,entry.S,hardirq.h,softirq.h)
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.6-pre1+fix(bitops,semaphore,softirq.h)
>     CqREEK	SolutionEngine
>     Bad
> 
> vanilla 2.4.5+fix(bitops,semaphore)
>     CqREEK	SolutionEngine
>     Bad
> --
> 
> 
> _______________________________________________
> linuxsh-dev mailing list
> lin...@li...
> http://lists.sourceforge.net/lists/listinfo/linuxsh-dev
> 
>-- End of excerpt from du...@se...