Re: [Intel-gfx] [PATCH] Revert "HAX iommu/intel: Ignore igfx_off"
On Thu, Sep 15, 2022 at 09:23:52AM +0100, Tvrtko Ursulin wrote: On 15/09/2022 09:09, Lucas De Marchi wrote: On Thu, Sep 15, 2022 at 08:53:16AM +0100, Tvrtko Ursulin wrote: On 14/09/2022 21:35, Lucas De Marchi wrote: This reverts commit 58f44e349cfc10a4f2208fd806829c8fd046480b. To be removed from the topic/core-for-CI branch. If CI's config is setting that, it just shouldn't do it. Looking at a random current CI execution, the command line is: Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 \ rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 \ softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 \ modprobe.blacklist=i915,snd_hda_intel ro So it should be safe to remove this commit. It's a patch which can silently bite so no complaints from me to remove it - just please upgrade the check from a "random current CI execution" to something stronger. Ack from CI folks would do. Well in the commit message I mentioned "To be removed from the topic/core-for-CI branch". This is how the topic/core-for-CI branch is handled until now. We simply remove the commit and force push. So there won't be any trace of commit messages for me to write anything stronger. To be at least registered in the mailing list what I used (no it was not this, but linking to the execution for this specific patch serves the purpose): https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_108575v1/bat-dg2-8/boot0.txt <5>[ 0.00] Linux version 6.0.0-rc5-Patchwork_108575v1-g37b0cd34584f+ (kbu...@ci-worker1.fi.intel.com) (gcc (Ubuntu 8.4.0-1ubuntu1~18.04) 8.4.0, GNU ld (GNU Binutils for Ubuntu) 2.30) #1 SMP PREEMPT_DYNAMIC Wed Sep 14 23:50:09 EEST 2022 <6>[ 0.00] Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 modprobe.blacklist=i915,snd_hda_intel ro Petri already acked this from the CI side via IRC today. That's all I suggested. :) Great. CI passed and this is acked by Petri and Tvrtko. Applying. thanks Lucas De Marchi Regards, Tvrtko
Re: [Intel-gfx] [PATCH] Revert "HAX iommu/intel: Ignore igfx_off"
On 15/09/2022 09:09, Lucas De Marchi wrote: On Thu, Sep 15, 2022 at 08:53:16AM +0100, Tvrtko Ursulin wrote: On 14/09/2022 21:35, Lucas De Marchi wrote: This reverts commit 58f44e349cfc10a4f2208fd806829c8fd046480b. To be removed from the topic/core-for-CI branch. If CI's config is setting that, it just shouldn't do it. Looking at a random current CI execution, the command line is: Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 \ rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 \ softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 \ modprobe.blacklist=i915,snd_hda_intel ro So it should be safe to remove this commit. It's a patch which can silently bite so no complaints from me to remove it - just please upgrade the check from a "random current CI execution" to something stronger. Ack from CI folks would do. Well in the commit message I mentioned "To be removed from the topic/core-for-CI branch". This is how the topic/core-for-CI branch is handled until now. We simply remove the commit and force push. So there won't be any trace of commit messages for me to write anything stronger. To be at least registered in the mailing list what I used (no it was not this, but linking to the execution for this specific patch serves the purpose): https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_108575v1/bat-dg2-8/boot0.txt <5>[ 0.00] Linux version 6.0.0-rc5-Patchwork_108575v1-g37b0cd34584f+ (kbu...@ci-worker1.fi.intel.com) (gcc (Ubuntu 8.4.0-1ubuntu1~18.04) 8.4.0, GNU ld (GNU Binutils for Ubuntu) 2.30) #1 SMP PREEMPT_DYNAMIC Wed Sep 14 23:50:09 EEST 2022 <6>[ 0.00] Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 modprobe.blacklist=i915,snd_hda_intel ro Petri already acked this from the CI side via IRC today. That's all I suggested. :) Regards, Tvrtko
Re: [Intel-gfx] [PATCH] Revert "HAX iommu/intel: Ignore igfx_off"
On Thu, Sep 15, 2022 at 08:53:16AM +0100, Tvrtko Ursulin wrote: On 14/09/2022 21:35, Lucas De Marchi wrote: This reverts commit 58f44e349cfc10a4f2208fd806829c8fd046480b. To be removed from the topic/core-for-CI branch. If CI's config is setting that, it just shouldn't do it. Looking at a random current CI execution, the command line is: Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 \ rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 \ softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 \ modprobe.blacklist=i915,snd_hda_intel ro So it should be safe to remove this commit. It's a patch which can silently bite so no complaints from me to remove it - just please upgrade the check from a "random current CI execution" to something stronger. Ack from CI folks would do. Well in the commit message I mentioned "To be removed from the topic/core-for-CI branch". This is how the topic/core-for-CI branch is handled until now. We simply remove the commit and force push. So there won't be any trace of commit messages for me to write anything stronger. To be at least registered in the mailing list what I used (no it was not this, but linking to the execution for this specific patch serves the purpose): https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_108575v1/bat-dg2-8/boot0.txt <5>[0.00] Linux version 6.0.0-rc5-Patchwork_108575v1-g37b0cd34584f+ (kbu...@ci-worker1.fi.intel.com) (gcc (Ubuntu 8.4.0-1ubuntu1~18.04) 8.4.0, GNU ld (GNU Binutils for Ubuntu) 2.30) #1 SMP PREEMPT_DYNAMIC Wed Sep 14 23:50:09 EEST 2022 <6>[0.00] Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 modprobe.blacklist=i915,snd_hda_intel ro Petri already acked this from the CI side via IRC today. thanks Lucas De Marchi
Re: [Intel-gfx] [PATCH] Revert "HAX iommu/intel: Ignore igfx_off"
On 14/09/2022 21:35, Lucas De Marchi wrote: This reverts commit 58f44e349cfc10a4f2208fd806829c8fd046480b. To be removed from the topic/core-for-CI branch. If CI's config is setting that, it just shouldn't do it. Looking at a random current CI execution, the command line is: Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 \ rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 \ softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 \ modprobe.blacklist=i915,snd_hda_intel ro So it should be safe to remove this commit. It's a patch which can silently bite so no complaints from me to remove it - just please upgrade the check from a "random current CI execution" to something stronger. Ack from CI folks would do. Regards, Tvrtko Signed-off-by: Lucas De Marchi --- drivers/iommu/intel/iommu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 598c9b86dfd8..1f2cd43cf9bc 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -323,6 +323,8 @@ static int __init intel_iommu_setup(char *str) no_platform_optin = 1; pr_info("IOMMU disabled\n"); } else if (!strncmp(str, "igfx_off", 8)) { + dmar_map_gfx = 0; + pr_info("Disable GFX device mapping\n"); } else if (!strncmp(str, "forcedac", 8)) { pr_warn("intel_iommu=forcedac deprecated; use iommu.forcedac instead\n"); iommu_dma_forcedac = true;
[Intel-gfx] [PATCH] Revert "HAX iommu/intel: Ignore igfx_off"
This reverts commit 58f44e349cfc10a4f2208fd806829c8fd046480b. To be removed from the topic/core-for-CI branch. If CI's config is setting that, it just shouldn't do it. Looking at a random current CI execution, the command line is: Command line: BOOT_IMAGE=/boot/drm_intel root=/dev/nvme0n1p2 \ rootwait fsck.repair=yes nmi_watchdog=panic,auto panic=5 \ softdog.soft_panic=5 drm.debug=0xe log_buf_len=1M trace_clock=global 3 \ modprobe.blacklist=i915,snd_hda_intel ro So it should be safe to remove this commit. Signed-off-by: Lucas De Marchi --- drivers/iommu/intel/iommu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 598c9b86dfd8..1f2cd43cf9bc 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -323,6 +323,8 @@ static int __init intel_iommu_setup(char *str) no_platform_optin = 1; pr_info("IOMMU disabled\n"); } else if (!strncmp(str, "igfx_off", 8)) { + dmar_map_gfx = 0; + pr_info("Disable GFX device mapping\n"); } else if (!strncmp(str, "forcedac", 8)) { pr_warn("intel_iommu=forcedac deprecated; use iommu.forcedac instead\n"); iommu_dma_forcedac = true; -- 2.37.3