Help needed fsck exited with signal 11
dirkw at rentec.com
Wed Nov 14 11:39:43 CST 2001
thx for the info. According to your webpage, I thought this is only
for "older" kernels. The message in red is between 2.4.9 and 2.4.10,
I have trouble with 2.4.14 and aacraid-20011105. What's the situation
here? And if CP's patch is applicable for this kernel, could somebody
please give me a pointer to it?
On Wed, 14 Nov 2001 Matt_Domsch at Dell.com wrote:
> > two questions: do you mean ">1GB" or ">=1GB"? secondly does it have
> > something to do with then fsck.ext2 hangs upon booting?
> It's actually "more than 896MB of RAM" to be strict. Above that, it's
> considered "High memory" (CONFIG_HIGHMEM4G or CONFIG_HIGHMEM64G). Problem
> is, the aacraid driver's scatter-gather functions during I/O are getting
> passed a buffer that resides in high memory, where on earlier kernels (pre
> 2.4.9-ac(really late)) this wasn't the case. Fortunately, Jens Axboe has
> been working on allowing the SCSI subsystem to make use of high memory
> properly, rather than use bounce buffers in low memory for I/O which really
> slows things down. That's the change that went in to 2.4.9-acX, and what's
> causing the breaking of the aacraid driver which wasn't expecting that. :-(
> The temporary workaround is to boot with "mem=896M". Not pretty, but it'll
> get you up-and-running.
> The permanent solution will be to apply Chris Pascoe's patch which allows
> I/O buffers to be in high memory (this was definitely a bug in the driver,
> though just now discovered), and while we're at it, try to root out all the
> other cases where this could break in the future.
> Matt Domsch
> Sr. Software Engineer
> Dell Linux Solutions
> #1 US Linux Server provider with 24% (IDC Sept 2001)
> #2 Worldwide Linux Server provider with 17% (IDC Sept 2001)
> #3 Unix provider with 18% in the US (Dataquest)!
Dirk Wetter @ Renaissance Techn.
mailto:<dirkw at rentec dot com>
More information about the Linux-PowerEdge