The NVIDIA 390 Driver Is Playing Nicely With Linux 4.15 Kernel

[phoronix]

Phoronix: The NVIDIA 390 Driver Is Playing Nicely With Linux 4.15 Kernel

For those NVIDIA Linux users reliant upon the proprietary driver and wanting to upgrade to the Linux 4.15 kernel that will be officially released within the next two weeks, the 390.12 driver is playing nicely...

http://www.phoronix.com/scan.php?page=news_item&px=Linux-4.15-NVIDIA-390-Driver

[kokoko3k]

Besides fixing the vRAM leakage

Is it really fixed? Do you run tests about it?
I'm asking because there is no mention in the changelog.

Here's the changelog.

Release highlights since 387.34:

• Added support for the following GPUs:
  • GeForce GTX 1050 Ti with Max-Q Design
  • GeForce GTX 1070 Ti
  • GeForce MX110
  • GeForce MX130
  • Quadro P500
  • TITAN V
  • TITAN Xp COLLECTORS EDITION
• Added new application profile settings, "EGLVisibleDGPUDevices" and "EGLVisibleTegraDevices", to control which discrete and Tegra GPU devices, respectively, may be enumerated by EGL. See the "Application Profiles" appendix of the driver README for more details.
• Corrected the SONAME of the copy of the libnvidia-egl-wayland library included in the .run installer package to libnvidia-egl-wayland.so.1. The SONAME had previously been versioned incorrectly with the full version number of the library.
• Updated nvidia.ko to veto the ACPI_VIDEO_NOTIFY_PROBE event on kernels that allow the handler for this event to be overridden, to improve interaction between the NVIDIA driver and acpi_video on display hotplug events.
• Updated the SLI Mosaic layout page in the nvidia-settings control panel to support topologies with up to 32 displays.
• Fixed a bug that prevented Xinerama Info from being handled properly in SLI or Base Mosaic layouts with more than 24 displays.
• Updated the X driver's composition pipeline (used for rotation, warp and blend, transformation matrices, etc) to also support stereo.
• Added an OpenGL stereo preview feature to the screen page in nvidia-settings.
• Fixed a bug where GetTexSubImage() would read incorrect data into a pixel buffer object when supplied with a target of GL_TEXTURE_1D_ARRAY and a non-zero yoffset value.
• Added support for generic active stereo with in-band DisplayPort signaling. The X configuration option "InbandStereoSignaling" is deprecated in favor of this stereo mode. See "Appendix B. X Config Options" in the README for more information.
• Modified the driver to avoid restoring framebuffer console modes on virtual reality head-mounted displays.
• Fixed a bug that could cause a system crash when using the new NVreg_EnableBacklightHandler kernel module parameter on GPUs with no displays connected

[theriddick]

Think it was meant to be fixed in previous version, which might be why its not mentioned this time around.

This is good news since I move to a 1080Ti Mini card in my system next week, probably the right decision too since Navi is now a 2019 product.

[sarmad]

are also safeguarded against Spectre vulnerability on NVIDIA's hardware

Wait, is nVidia hardware itself vulnerable to Spectre, or is this a fix for the driver itself since the driver itself runs on the CPU, not the GPU?

[Michael]

Wait, is nVidia hardware itself vulnerable to Spectre, or is this a fix for the driver itself since the driver itself runs on the CPU, not the GPU?

NVIDIA has confirmed their hardware is vulnerable.

[sarmad]

NVIDIA has confirmed their hardware is vulnerable.

Wow!!

[Guest]

NVIDIA has confirmed their hardware is vulnerable.

Wonder who confirmed that, obviously their hardware wasn't vulnerable, their drivers were.
Funnily enough they explicitly updated CVE post to say so.


Edit: Actually, I can't make the claim that their hw isn't vulnerable, nobody has done research to confirm/disprove that AFAIK.
Wonder if CL/Cuda kernels can be affected.

Edit2: https://web.archive.org/web/20180107...tail/a_id/4611

[Michael]

Wonder who confirmed that, obviously their hardware isn't vulnerable, their drivers are.
Funnily enough they explicitly updated CVE post to say so.


Edit: Actually, I can't make the claim that their hw isn't vulnerable, nobody has done research to confirm/disprove that AFAIK.
Wonder if CL/Cuda kernels can be affected.

sarmad

Yeah they clarified now they are talking about driver safeguards.

[Guest]

Even in the original revision they talk about "GPU Display Driver" being potentially affected, I suspect that somebody misinterpreted it as "nvidia being affected" and hence why they distanced even further (see the side-by-side comparison of new/archive.org versions). Could this be that somebody misinterpreted your original article, wrote a shitty fluff piece on routers/nbc/reddit and suddenly we have bunch of new articles feeding on the drama?

[sarmad]

sarmad

Yeah they clarified now they are talking about driver safeguards.

Then it might be worth updating your article to clarify that.