loader(8) readin failed on 7.2R and later including 8.0R

[John Baldwin]

On Thursday 18 March 2010 7:10:59 am Carsten Bäcker wrote:
> Hello,
> 
> i ran into a similar problem when recently trying to install 8.0R on an
> embedded system. We've been using this hardware for a couple of
> years with FreeBSD 4.11 and didn't see any problem like this.
> 
> Disabling the memory hole (15-16M) solved the problem here, but
> that shouldn't be a solution. I wonder whether it's a loader-, or a
> BIOS-problem, since memory-allocation should respect reserved
> areas.

The problem is likely because an 8.0 kernel is simply larger than a 4.11
kernel.  4.11's loader would have had the same issue.  The problem (as it
were), is that we expect to be able to load the kernel + modules into one
contiguous chunk of RAM, starting at 4MB (PAE and amd64 kernels start at
2MB).
 
> Best regards
> Carsten Bäcker
> 
> 
> On Sunday 06 December 2009 12:16:36 am Hiroki Sato wrote:
> >  Hiroki Sato<h... at freebsd.org>  wrote
> >    in<20091205.184250.201700943.... at allbsd.org>:
> >
> >  hr>   A summary so far is:
> >  hr>
> >  hr>   1)  a<8MB 7.1R kernel + stock 8.0R loader
> >  hr>   2a) a>8MB 8.0R kernel + stock 8.0R loader
> >  hr>   2b) a>8MB 8.0R kernel + 8.0R loader with LOADER_NO_GPT_SUPPORT=yes
> >  hr>   2c) a>8MB 8.0R kernel + loader with your patch
> >  hr>   3a) a<8MB 8.0R kernel + stock 8.0R loader
> >  hr>   3b) a<8MB 8.0R kernel + 8.0R loader with LOADER_NO_GPT_SUPPORT=yes
> >  hr>   3c) a<8MB 8.0R kernel + loader with your patch
> >
> >   Grr, I double-checked how it got stuck, then I found the console
> >   redirect was disabled because of an old device.hints.  The revised
> >   summary is:
> >
> >         loading text      loading syms       boot
> >    1)   OK                OK                 OK
> >    2a)  "readin failed"   -                  -
> >    2b)  OK                "skipped!"         OK
> >    2c)  OK                "skipped!"         OK
> >    3a)  OK                OK                 OK
> >    3b)  OK                OK                 OK
> >    3c)  OK                OK                 OK
> >
> >   So, the case 2c shows that your patch solves the problem in the case
> >   2a.  Thank you! :)
> >
> >   Loading>8MB kernel works now, but loading syms sections still fails
> >   even in the case 2c.
> 
> Ok.  Your system's SMAP is kind of weird (it has a very small region above
> 1MB, so it may not deal well with "large" kernels, though I thought it had
> enough room for at least a 12MB kernel.  Hmm, the size of the kernel file may
> be deceptive though since it does not include BSS.  I wonder if it is trying
> to load the symbols after the BSS.

-- 
John Baldwin