<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> <meta name="Generator" content="Microsoft Exchange Server">  <style></style> </head> <body> <font face="Calibri, sans-serif" size="2"> <div style="text-align: justify; ">SRU Justification:</div> <div style="text-align: justify; ">These two patches fix one important bug (<a href="https://bugs.freedesktop.org/show_bug.cgi?id=27108"><font color="#0000FF"><u>https://bugs.freedesktop.org/show_bug.cgi?id=27108</u></font></a>) about interrupt handling where reporters said “On some machines the Xserver freezes inside the intel driver in a ioctl (all output ceases) until the mouse is moved.”</div> <div style="text-align: justify; ">And from comments, this issue has only been seen on Ironlake gfx so far.</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; ">Test Cases:</div> <div style="text-align: justify; ">From bug reporter’s comments:</div> <div style="text-align: justify; ">Triggering the bug is extremely difficult, and seems to depend on weird</div> <div style="text-align: justify; ">circumstances (I considered air pressure or state of the moon for a while). We</div> <div style="text-align: justify; ">had one machine shipped to us that exposed the bug, when it arrived (it even</div> <div style="text-align: justify; ">woke up from suspend, so it was in exactly the same state) we weren't able to</div> <div style="text-align: justify; ">reproduce even on this machine. But colleagues we trust have seen the bug with</div> <div style="text-align: justify; ">their very eyes.</div> <div style="text-align: justify; "><font face="Times New Roman, serif"> </font></div> <div style="text-align: justify; ">Thanks,</div> <div style="text-align: justify; ">Yingying</div> <div style="text-align: justify; "><font face="Times New Roman, serif"> </font></div> <div style="text-align: justify; ">---</div> <div style="text-align: justify; ">commit e552eb7038a36d9b18860f525aa02875e313fe16</div> <div style="text-align: justify; ">Author: Jesse Barnes <jbarnes@virtuousgeek.org></div> <div style="text-align: justify; ">Date: Wed Apr 21 11:39:23 2010 -0700</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> drm/i915: use PIPE_CONTROL instruction on Ironlake and Sandy Bridge</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> Since 965, the hardware has supported the PIPE_CONTROL command, which</div> <div style="text-align: justify; "> provides fine grained GPU cache flushing control. On recent chipsets,</div> <div style="text-align: justify; "> this instruction is required for reliable interrupt and sequence number</div> <div style="text-align: justify; "> reporting in the driver.</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> So add support for this instruction, including workarounds, on Ironlake</div> <div style="text-align: justify; "> and Sandy Bridge hardware.</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27108"> https://bugs.freedesktop.org/show_bug.cgi?id=27108</a></div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org></div> <div style="text-align: justify; "> Tested-by: Chris Wilson <chris@chris-wilson.co.uk></div> <div style="text-align: justify; "> Signed-off-by: Eric Anholt <eric@anholt.net></div> <div style="text-align: justify; "> </div> <div style="text-align: justify; ">commit 1918ad77f7f908ed67cf37c505c6ad4ac52f1ecf</div> <div style="text-align: justify; ">Author: Jesse Barnes <jbarnes@virtuousgeek.org></div> <div style="text-align: justify; ">Date: Fri Apr 23 09:32:23 2010 -0700</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> drm/i915: fix non-Ironlake 965 class crashes</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> My PIPE_CONTROL fix (just sent via Eric's tree) was buggy; I was</div> <div style="text-align: justify; "> testing a whole set of patches together and missed a conversion to the</div> <div style="text-align: justify; "> new HAS_PIPE_CONTROL macro, which will cause breakage on non-Ironlake</div> <div style="text-align: justify; "> 965 class chips. Fortunately, the fix is trivial and has been tested.</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> Be sure to use the HAS_PIPE_CONTROL macro in i915_get_gem_seqno, or</div> <div style="text-align: justify; "> we'll end up reading the wrong graphics memory, likely causing hangs,</div> <div style="text-align: justify; "> crashes, or worse.</div> <div style="text-align: justify; "> </div> <div style="text-align: justify; "> Reported-by: Zdenek Kabelac <zdenek.kabelac@gmail.com></div> <div style="text-align: justify; "> Reported-by: Toralf Förster <toralf.foerster@gmx.de></div> <div style="text-align: justify; "> Tested-by: Toralf Förster <toralf.foerster@gmx.de></div> <div style="text-align: justify; "> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org></div> <div style="text-align: justify; "> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org></div> <div style="text-align: justify; "><font face="Times New Roman, serif"> </font></div> <div style="text-align: justify; "><font face="Times New Roman, serif"> </font></div> </font> </body> </html>