Tested the Intel e1000e driver in Kernel 5.4.1, and this still fails with some Ethernet NICs, but is OK on others.. On my HP Z820, there are two (different) Intel NICs, and the e1000e driver works with one, but fails with the other: LAN connection to eth0 (Intel 82579LM) fails: [ 125.292581] userif-2: sent link down event. … Continue reading →
After updating my main HP Z820 system to Kernel 5.4, I found that the previous Intel e1000e Ethernet problem I had encountered on my test system with Kernel 5.4-rc6 seemed to have reappeared. (see https://rglinuxtech.com/?p=2632 ). But – this time, two of my systems with e1000e drivers were OK, but the third – the HP Z820 – … Continue reading →
Kernel 5.4 has been released, and brief details of changes since -rc8 are here: http://lkml.iu.edu/hypermail/linux/kernel/1911.3/00186.html As expected, the latest NVIDIA (440.36) and patched VMware 15.5.1 both compile and load/run OK. Robert Gadsdon. November 24th, 2019
Continue reading →Kernel 5.4-rc8 is out, and brief details are here: http://lkml.iu.edu/hypermail/linux/kernel/1911.2/01078.html The latest NVIDIA (440.31), and patched VMware 15.5.1 compile and load/run OK. The Intel e1000e Ethernet driver had been broken since -rc1, but this has finally been fixed, and appears to operate correctly, now. Robert Gadsdon. November 17th 2019.
Continue reading →VMware have released version 15.5.1, and details are here: https://docs.vmware.com/en/VMware-Workstation-Pro/15.5/rn/VMware-Workstation-1551-Pro-Release-Notes.html The Linux-kernel-related stuff appears to be unchanged, and compiles (with several warnings..) with Kernel 5.3 (tested with 5.3.11). I got some odd behaviour (shared folders becoming disabled?) when running this unpatched version, and so applied the patched 15.5.0 vmmon/vmnet, which seems to work better.. Robert Gadsdon. November … Continue reading →
NVIDIA have just released driver 440.31, and details are here: https://devtalk.nvidia.com/default/topic/1065987/announcements-and-news/linux-solaris-and-freebsd-driver-440-31-long-lived-branch-release-/ I have tested this with Kernel 5.4-rc6, and – as mentioned in the release notes – there is no annoying ‘Module.symvers…‘ message any more. Robert Gadsdon. November 4th, 2019.
Continue reading →Kernel 5.4-rc3 has been released, and brief details of changes since -rc2 are here: http://lkml.iu.edu/hypermail/linux/kernel/1910.1/05411.html The patched versions of VMware 15.5.0 and NVIDIA 435.21 now compile without all the spurious warnings etc.. Robert Gadsdon. October 13th 2019.
Continue reading →Thanks to loqs, there is a patch for driver 435.21 and kernel 5.4-rc, at http://ix.io/1MRE . I have tested this with 5.4-rc2, and it now compiles successfully: ………………………………. WARNING: module nvidia uses symbol filp_open from namespace /NVIDIA-Linux-x86_64-435.21-54rc2/kernel/nvidia-drm/nvidia-drm-gem-user-memory.o: $(deps_/home/rgadsdon/kernel/NVIDIA-Linux-x86_64-435.21-54rc2/kernel/nvidia-drm/nvidia-drm-gem-user-memory.o), but does not import it. WARNING: module nvidia uses symbol cpufreq_get from namespace VIDIA-Linux-x86_64-435.21-54rc2/kernel/nvidia-drm/nvidia-drm-gem-user-memory.o: $(deps_/home/rgadsdon/kernel/NVIDIA-Linux-x86_64-435.21-54rc2/kernel/nvidia-drm/nvidia-drm-gem-user-memory.o), but does not import … Continue reading →
Kernel 5.4-rc2 has been released, and brief details are here: http://lkml.iu.edu/hypermail/linux/kernel/1910.0/07213.html The odd behaviour of VMware and NVIDIA compiles still occurs, but there are patches submitted (see comments under -rc1 article https://rglinuxtech.com/?p=2614 for more details). So far, I have not seen a fix for the NVIDIA compile failure. Robert Gadsdon. October 7th 2019.
Continue reading →Kernel 5.4-rc1 is out – a day later than usual, and brief details are here: http://lkml.iu.edu/hypermail/linux/kernel/1909.3/04895.html My tests with VMware and NVIDIA revealed some odd errors/warning relating to kernel symbols.. With the latest NVIDIA (435.21) I got the following: # make make[1]: Entering directory ‘/usr/src/linux-5.4-rc1’ The Module.symvers file is missing, or does not contain any symbols exported … Continue reading →