similar to: [RFC PATCH v2] clk: Remove BYPASS_PLL_CHECK from PLLs

I have looked at problem with Fermi GPUs where changing to higher clock led to really bad perfomance (with GpuTest 20x worse perfomance) and later also crashes of the nouveau. It seemed to be affected by Shader Clock in Voltage Entries in the video BIOS. Disabling BYPASS_PLL_CHECK in CLK0_CTRL seems to completely fix the issue. I have tried to search this BYPASS_PLL_CHECK in Nvidia traces but

I have looked at problem with Fermi GPUs where changing to higher clock led to really bad perfomance (with GpuTest 20x worse perfomance) and later also crashes of the nouveau. It seemed to be affected by Shader Clock in Voltage Entries in the video BIOS. Disabling BYPASS_PLL_CHECK in CLK0_CTRL seems to completely fix the issue. I have tried to search this BYPASS_PLL_CHECK in Nvidia traces but

I have looked at problem with Fermi GPUs where changing to higher clock led to really bad perfomance (with GpuTest 20x worse perfomance) and later also crashes of the nouveau. It seemed to be affected by Shader Clock in Voltage Entries in the video BIOS. Disabling BYPASS_PLL_CHECK in CLK0_CTRL seems to completely fix the issue. I have tried to search this BYPASS_PLL_CHECK in Nvidia traces but

https://bugs.freedesktop.org/show_bug.cgi?id=111724 Bug ID: 111724 Summary: NVE6 (GK106) memory re-clocking breaks GpuTest plot3d benchmark Product: xorg Version: unspecified Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: normal Priority: not set

trello card: https://trello.com/c/admzDRvd/50-reduce-performance-refuse-to-boot-if-not-all-the-power-connectors-are-plugged Mark Menzynski (4): moved gpio so it is sorted by values gpio: checking if gpu power cable is connected gpio: added power check for another GPIO gpio: added power check for another GPIO drm/nouveau/include/nvkm/subdev/bios/gpio.h | 5 ++++-

Trello: https://trello.com/c/admzDRvd/50-reduce-performance-refuse-to-boot-if-not-all-the-power-connectors-are-plugged Fixed commit messages. Mark Menzynski (4): bios/gpio: sort gpios by values gpio: fail if gpu external power is missing gpio: check the gpio function 16 in the power check as well gpio: check function 76 in the power check as well

Added config override for power checks. Mark Menzynski (4): bios/gpio: sort gpios by values gpio: fail if gpu external power is missing gpio: check the gpio function 16 in the power check as well gpio: check function 76 in the power check as well drm/nouveau/include/nvkm/subdev/bios/gpio.h | 5 +++- drm/nouveau/nvkm/subdev/gpio/base.c | 32 +++++++++++++++++++++ 2 files

Please add a config override to skip this, since we'll invariably get it wrong for some setup, and should be able to provide users with workarounds while the issue is being worked out. On Mon, Jul 15, 2019 at 5:43 AM Mark Menzynski <mmenzyns at redhat.com> wrote: > > Currently, nouveau doesn't check if GPU is missing power. This > patch makes nouveau fail when this happens

Some, mostly Fermi, vbioses appear to have zero max voltage. That causes Nouveau to not parse voltage entries, thus users not being able to set higher clocks. When changing this value Nvidia driver still appeared to ignore it, and I wasn't able to find out why, thus the code is ignoring the value if it is zero. CC: Maarten Lankhorst <maarten.lankhorst at linux.intel.com> Signed-off-by:

Corresponds with GT215. Don't rely on the lock test logic being unconditionally enabled, and disable test logic when done (presumably to save power). Warn when locking fails. Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c | 15 ++++++++++++--- drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c | 15 ++++++++++++--- 2 files changed, 24

Corresponds with GT215. Don't rely on the lock test logic being unconditionally enabled, and disable test logic when done (presumably to save power). v2: Remove warning, nvkm_msec already warns on time-out Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c | 8 +++++++- drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c | 8 +++++++- 2

Following a series of three patches, two of which have been sitting in my tree for a while, the third is the result of some inspection of an NV134 BIOS that seems to use the 0xaf upcode to upload training patterns. Please test! Roy Ps. Sorry they come from yet another e-mail address. My previous provider, eclipso, actively blocks users of git send-email. Inquiries fall on deaf ears, hence I

From: Mark Menzynski <mmenzyns at redhat.com> [ Upstream commit a1af2afbd244089560794c260b2d4326a86e39b6 ] Some, mostly Fermi, vbioses appear to have zero max voltage. That causes Nouveau to not parse voltage entries, thus users not being able to set higher clocks. When changing this value Nvidia driver still appeared to ignore it, and I wasn't able to find out why, thus the code is

From: Mark Menzynski <mmenzyns at redhat.com> [ Upstream commit a1af2afbd244089560794c260b2d4326a86e39b6 ] Some, mostly Fermi, vbioses appear to have zero max voltage. That causes Nouveau to not parse voltage entries, thus users not being able to set higher clocks. When changing this value Nvidia driver still appeared to ignore it, and I wasn't able to find out why, thus the code is

From: Mark Menzynski <mmenzyns at redhat.com> [ Upstream commit a1af2afbd244089560794c260b2d4326a86e39b6 ] Some, mostly Fermi, vbioses appear to have zero max voltage. That causes Nouveau to not parse voltage entries, thus users not being able to set higher clocks. When changing this value Nvidia driver still appeared to ignore it, and I wasn't able to find out why, thus the code is

From: Mark Menzynski <mmenzyns at redhat.com> [ Upstream commit a1af2afbd244089560794c260b2d4326a86e39b6 ] Some, mostly Fermi, vbioses appear to have zero max voltage. That causes Nouveau to not parse voltage entries, thus users not being able to set higher clocks. When changing this value Nvidia driver still appeared to ignore it, and I wasn't able to find out why, thus the code is

On Mon, 15 Jul 2019 at 22:26, Ilia Mirkin <imirkin at alum.mit.edu> wrote: > > Please add a config override to skip this, since we'll invariably get > it wrong for some setup, and should be able to provide users with > workarounds while the issue is being worked out. Yeah, this makes me nervous as well. In the very least, I'd like a config option, but I'm still

Signed-off-by: Mark Menzynski <mmenzyns at redhat.com> --- drm/nouveau/include/nvkm/subdev/bios/gpio.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drm/nouveau/include/nvkm/subdev/bios/gpio.h b/drm/nouveau/include/nvkm/subdev/bios/gpio.h index b71a3555..2f40935f 100644 --- a/drm/nouveau/include/nvkm/subdev/bios/gpio.h +++ b/drm/nouveau/include/nvkm/subdev/bios/gpio.h

Tested on TU104, GP106 and GF100 Signed-off-by: Mark Menzynski <mmenzyns at redhat.com> --- drm/nouveau/include/nvkm/subdev/bios/gpio.h | 1 + drm/nouveau/nvkm/subdev/gpio/base.c | 23 +++++++++++++++++++++ 2 files changed, 24 insertions(+) diff --git a/drm/nouveau/include/nvkm/subdev/bios/gpio.h b/drm/nouveau/include/nvkm/subdev/bios/gpio.h index 2f40935f..a70ec9e8 100644 ---

this one is mainly used on Rankine and Curie and rarely on Tesla GPUs untested, wrote according to documentation Signed-off-by: Mark Menzynski <mmenzyns at redhat.com> --- drm/nouveau/include/nvkm/subdev/bios/gpio.h | 1 + drm/nouveau/nvkm/subdev/gpio/base.c | 1 + 2 files changed, 2 insertions(+) diff --git a/drm/nouveau/include/nvkm/subdev/bios/gpio.h