Re: port-i386/42966: kernel built from today's netbsd-5 crashes

[Anne Bennett <anne%porcupine.montreal.qc.ca@localhost>]

The following reply was made to PR port-i386/42966; it has been noted by GNATS.

From: Anne Bennett <anne%porcupine.montreal.qc.ca@localhost>
To: matthew green <mrg%eterna.com.au@localhost>
Cc: port-i386-maintainer%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
    netbsd-bugs%netbsd.org@localhost, gnats-bugs%netbsd.org@localhost,
    "David H. Gutteridge" <dhgutteridge%sympatico.ca@localhost>
Subject: Re: port-i386/42966: kernel built from today's netbsd-5 crashes
Date: Wed, 17 Mar 2010 23:40:10 -0400

 Matthew,
 
 Thanks for looking into this.
 
 > it comes from this call in i915_dma.c:
 > 
 >      drm_pci_alloc(dev, PAGE_SIZE, PAGE_SIZE, 0xffffffff);
 > 
 > and the final argument is passed as the boundary.  notice how it is
 > not a power of two?  it really wants to be 0x1.0000.0000.
 > 
 > anne, david, can one of you try changing this to 0?  or if you have
 > 2GB of ram or less, try 0x80000000...
 
 Kernel says:
   total memory = 2038 MB
   avail memory = 1991 MB
 
 Any preference as to whether I try 0x80000000 or 0x0 (to replace
 0xffffffff, I assume)?  Which one are you hoping to patch to in
 the end?
 
 Meanwhile, I have two files "i915_dma.c" with different contents:
    (1) /usr/src/sys/dev/pci/drm/i915_dma.c
    (2) /usr/src/sys/external/bsd/drm/dist/shared-core/i915_dma.c
 
 and both have statements like the one above:
 
 (1)
         /* Program Hardware Status Page */
         dev_priv->status_page_dmah = drm_pci_alloc(dev, PAGE_SIZE, PAGE_SIZE,
             0xffffffff);
 
 (2)
         /* Program Hardware Status Page */
         DRM_UNLOCK();
         dev_priv->status_page_dmah =
                 drm_pci_alloc(dev, PAGE_SIZE, PAGE_SIZE, 0xffffffff);
         DRM_LOCK();
 
 ... do I change both?
 
 Anne.
 -- 
 Ms. Anne Bennett, as a private citizen:  
anne%porcupine.montreal.qc.ca@localhost
 Also reachable more officially at work:  anne%encs.concordia.ca@localhost