| Message ID | 1428689711-3541-5-git-send-email-rodrigo.vivi@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
I'm seeing the same behaviour with this patchset. After boot, X works fine but I get a rolling display on fbcon (the contents appear to be moving horizontally very quickly around the middle of the screen). If the screen is turned off and on again, X now only updates the screen once every second or so but fbcon works. If I suspend and resume, things go back to the working state until the next screen power cycle.
Tested-By: Intel Graphics QA PRTS (Patch Regression Test System Contact: shuang.he@intel.com)
Task id: 6180
-------------------------------------Summary-------------------------------------
Platform Delta drm-intel-nightly Series Applied
PNV -5 276/276 271/276
ILK 301/301 301/301
SNB 316/316 316/316
IVB 328/328 328/328
BYT 285/285 285/285
HSW 394/394 394/394
BDW 321/321 321/321
-------------------------------------Detailed-------------------------------------
Platform Test drm-intel-nightly Series Applied
*PNV igt@gem_fence_thrash@bo-write-verify-none PASS(2) FAIL(1)PASS(1)
*PNV igt@gem_fence_thrash@bo-write-verify-threaded-none PASS(2) FAIL(1)PASS(1)
*PNV igt@gem_fence_thrash@bo-write-verify-x PASS(2) FAIL(1)PASS(1)
*PNV igt@gem_fence_thrash@bo-write-verify-y PASS(2) FAIL(1)PASS(1)
PNV igt@gen3_render_mixed_blits FAIL(4)PASS(4) FAIL(1)PASS(1)
Note: You need to pay more attention to line start with '*'
Hi Matthew, Could you please check if you can reproduce your issue using this branch: http://cgit.freedesktop.org/~vivijim/drm-intel/log/?h=intel_psr Also, 1. What Platform are you using? BDW? 2. What desktop environment do you use? 3. What is your resolution? 4. Is IPS running? 5. Could you please grab dmesg with drm.debug=0xe? 6. Could you please paste a sequence of cat /sys/kernel/debug/dri/0/i915_edp_psr_status when facing the fbcon and X issues? Thanks in advance, Rodrigo. On Fri, Apr 10, 2015 at 3:05 PM, Matthew Garrett <mjg59@srcf.ucam.org> wrote: > I'm seeing the same behaviour with this patchset. After boot, X works > fine but I get a rolling display on fbcon (the contents appear to be > moving horizontally very quickly around the middle of the screen). If > the screen is turned off and on again, X now only updates the screen > once every second or so but fbcon works. If I suspend and resume, things > go back to the working state until the next screen power cycle. > > -- > Matthew Garrett | mjg59@srcf.ucam.org > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Another questions, Are you using powertop --auto-tune? If so, can you please try to repdoruce X slowness issue on these 2 scenarios: 1. without doing the powertop auto-tune and psr enabled. 2. with powertop auto-tune but with PSR disabled by i915.enable_psr=0 Thanks in advance, Rodrigo. On Mon, Apr 13, 2015 at 4:09 PM, Rodrigo Vivi <rodrigo.vivi@gmail.com> wrote: > Hi Matthew, > > Could you please check if you can reproduce your issue using this > branch: http://cgit.freedesktop.org/~vivijim/drm-intel/log/?h=intel_psr > > Also, > 1. What Platform are you using? BDW? > 2. What desktop environment do you use? > 3. What is your resolution? > 4. Is IPS running? > 5. Could you please grab dmesg with drm.debug=0xe? > 6. Could you please paste a sequence of cat > /sys/kernel/debug/dri/0/i915_edp_psr_status > when facing the fbcon and X issues? > > Thanks in advance, > Rodrigo. > > On Fri, Apr 10, 2015 at 3:05 PM, Matthew Garrett <mjg59@srcf.ucam.org> wrote: >> I'm seeing the same behaviour with this patchset. After boot, X works >> fine but I get a rolling display on fbcon (the contents appear to be >> moving horizontally very quickly around the middle of the screen). If >> the screen is turned off and on again, X now only updates the screen >> once every second or so but fbcon works. If I suspend and resume, things >> go back to the working state until the next screen power cycle. >> >> -- >> Matthew Garrett | mjg59@srcf.ucam.org >> _______________________________________________ >> Intel-gfx mailing list >> Intel-gfx@lists.freedesktop.org >> http://lists.freedesktop.org/mailman/listinfo/intel-gfx > > > > -- > Rodrigo Vivi > Blog: http://blog.vivi.eng.br
On Mon, Apr 13, 2015 at 04:46:29PM -0700, Rodrigo Vivi wrote: > Another questions, > > Are you using powertop --auto-tune? > > If so, can you please try to repdoruce X slowness issue on these 2 scenarios: > 1. without doing the powertop auto-tune and psr enabled. Ah! Yes, this is the problem. If runtime PM is enabled on the i915 PCI device (and the HDMI HDA device), things break. If it's disabled, everything works fine. I hope that helps narrow it down!
On Sat, 2015-04-18 at 08:27 +0100, Matthew Garrett wrote: > On Mon, Apr 13, 2015 at 04:46:29PM -0700, Rodrigo Vivi wrote: > > Another questions, > > > > Are you using powertop --auto-tune? > > > > If so, can you please try to repdoruce X slowness issue on these 2 scenarios: > > 1. without doing the powertop auto-tune and psr enabled. > > Ah! Yes, this is the problem. If runtime PM is enabled on the i915 PCI > device (and the HDMI HDA device), things break. If it's disabled, > everything works fine. I hope that helps narrow it down! > Are you using external USB keyboard and mouse? I can just face this slowness when using USB, if I don't use any USB everything is fine. If switch all USB powertop scripts to "Bad" I also can use my system reliably. This seems a bug in usb power management. Could you please verify if this is the same that I'm facing here? One extra thing, could you confirm that you face this behaviour even with i915.enable_psr=0 so Daniel can accept this last patch? Thank you very much, Rodrigo.
Hi Matthew, here is the patch I've mentioned on irc today: http://cgit.freedesktop.org/~vivijim/drm-intel/commit/?h=psr_for_mjg59&id=83809492138f2395bfb12c19e6de916de64b9246 And I prepared this branch for now: http://cgit.freedesktop.org/~vivijim/drm-intel/log/?h=psr_for_mjg59 I'm not sending yet the patches because I still face the missed screen during boot that you had mentioned. As soon as I fixed it I'll submit everything. Thanks, Rodrigo. On Mon, Apr 20, 2015 at 7:43 AM, Vivi, Rodrigo <rodrigo.vivi@intel.com> wrote: > On Sat, 2015-04-18 at 08:27 +0100, Matthew Garrett wrote: >> On Mon, Apr 13, 2015 at 04:46:29PM -0700, Rodrigo Vivi wrote: >> > Another questions, >> > >> > Are you using powertop --auto-tune? >> > >> > If so, can you please try to repdoruce X slowness issue on these 2 scenarios: >> > 1. without doing the powertop auto-tune and psr enabled. >> >> Ah! Yes, this is the problem. If runtime PM is enabled on the i915 PCI >> device (and the HDMI HDA device), things break. If it's disabled, >> everything works fine. I hope that helps narrow it down! >> > > Are you using external USB keyboard and mouse? I can just face this > slowness when using USB, if I don't use any USB everything is fine. If > switch all USB powertop scripts to "Bad" I also can use my system > reliably. This seems a bug in usb power management. Could you please > verify if this is the same that I'm facing here? > > One extra thing, could you confirm that you face this behaviour even > with i915.enable_psr=0 so Daniel can accept this last patch? > > Thank you very much, > Rodrigo.
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index bb64415..935951f 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -37,7 +37,7 @@ struct i915_params i915 __read_mostly = { .enable_execlists = -1, .enable_hangcheck = true, .enable_ppgtt = -1, - .enable_psr = 0, + .enable_psr = 1, .preliminary_hw_support = IS_ENABLED(CONFIG_DRM_I915_PRELIMINARY_HW_SUPPORT), .disable_power_well = 1, .enable_ips = 1, @@ -123,7 +123,7 @@ MODULE_PARM_DESC(enable_execlists, "(-1=auto [default], 0=disabled, 1=enabled)"); module_param_named(enable_psr, i915.enable_psr, int, 0600); -MODULE_PARM_DESC(enable_psr, "Enable PSR (default: false)"); +MODULE_PARM_DESC(enable_psr, "Enable PSR (default: true)"); module_param_named(preliminary_hw_support, i915.preliminary_hw_support, int, 0600); MODULE_PARM_DESC(preliminary_hw_support,
With a reliable frontbuffer tracking and all instability corner cases solved let's re-enabled PSR by default on all supported platforms. Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> --- drivers/gpu/drm/i915/i915_params.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)