Hi, everyone I noticed the nouveau module dropped support for the msi parameter on Linux 3.7. I've been enabling message signaled interrupts on my ION/MCP79 for a while with no adverse effects. Is this intended behaviour our just fallout from the rewrite? (I'm not subscribed to this mailing list.) Thanks, Rui Salvaterra -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.freedesktop.org/archives/nouveau/attachments/20121110/670913a0/attachment.html>
On 10 November 2012 19:31, Rui Salvaterra <rsalvaterra at gmail.com> wrote:> I noticed the nouveau module dropped support for the msi parameter on Linux > 3.7. I've been enabling message signaled interrupts on my ION/MCP79 for a > while with no adverse effects. Is this intended behaviour our just fallout > from the rewrite? (I'm not subscribed to this mailing list.)I was going to ask exactly the same question. For some platforms IOAPICs (thus legacy interrupts) aren't possible eg with PCIe hotplug or such, so MSI interrupts are required. The PCIe spec mandates that they are available for perhaps these purposes, just there is no mechanism in linux to test and fall back to legacy interrupts if there are platform issues. Thanks, Daniel -- Daniel J Blueman
Possibly Parallel Threads
- [Bug 104672] New: [NVAC] Compiz segfaults on login on Ubuntu 17.10 with Unity 7 (kernel DRM bug?)
- [Bug 54786] New: random crashes of X with failed to idle channel 0xcccc0000
- [Bug 57350] [nouveau, linux-3.7-rc] Broken cursor and kernel log swamped with trapped reads/writes from BAR/PFIFO_READ/FB
- Nouveau failing during probe followed by GPF on 3.13-rc2
- [PATCH] drm/nouveau: fix bug id typo in comment