Re: Bug#687136: update

[Jim McCloskey <mcclosk@ucsc.edu>]

* Jim McCloskey (mcclosk@ucsc.edu) wrote:

  |>  This is what I've done so far to try to sort this bug out:
  |>  
  |>   . tried different monitors
  |>   . tested the memory
  |>   . upgraded the BIOS
  |>   . tried the card in a different machine
  |>   . tried a self-compiled version of (Debianized) kernel 3.4
  |>   . tried libdrm packages built myself from:
  |>  
  |>     git://git.debian.org/git/pkg-xorg/lib/libdrm --branch debian-experimental    
  |>     (September 25th)
  |>  
  |>  No luck. I get X freezes in every condition except when the firmware package is
  |>  uninstalled and no acceleration is attempted. 
  |>  
  |>  Presumably the most relevant sign of trouble is the fact that GPU lockups
  |>  are reported:
  |>  
  |>    Oct  5 14:00:17 ohlone kernel: [   34.436052] GPU lockup (waiting for 0x00000004 last fence id 0x00000001)
  |>    Oct  5 14:00:17 ohlone kernel: [   34.437223] radeon 0000:01:00.0: GPU softreset
  |>    Oct  5 14:00:17 ohlone kernel: [   34.436052] GPU lockup (waiting for 0x00000004 last fence id 0x00000001)
  |>    Oct  5 14:00:17 ohlone kernel: [   34.437223] radeon 0000:01:00.0: GPU softreset
  |>     
  |>  at intervals of 12 to 20 seconds.
  |>  
  |>  At this point I think I have to just give up,

I resorted then to using the proprietary driver from AMD.

However, two days ago I upgraded from wheezy to jessie and installed kernel version
3.17-1. I purged fglrx and now rely again on the radeon driver. The problem reported
above is gone and all works very well (including acceleration). It's hard for me to know
which aspect of the upgrade was most important in resolving this, but in any case I'm
very happy indeed with the outcome and I suggest that the bug be closed at this point.

Thanks everyone,

Jim