[Bug 285492] NVIDIA Connect X-7 NIC:The mlx5(3.7.1) driver that comes along with 14.2 detects MFA1A00 cable as a 100G-KR4 cable incorrectly
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=285492 Bug ID: 285492 Summary: NVIDIA Connect X-7 NIC:The mlx5(3.7.1) driver that comes along with 14.2 detects MFA1A00 cable as a 100G-KR4 cable incorrectly Product: Base System Version: 14.2-STABLE Hardware: amd64 OS: Any Status: New Severity: Affects Some People Priority: --- Component: kern Assignee: b...@freebsd.org Reporter: nirmalanand.jebaku...@gmail.com Created attachment 258755 --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=258755&action=edit Cable specification I am running FreeBSD 14.2 with a NVIDIA Connect X-7 NIC. The mlx5(3.7.1) driver that comes along with the base kernel image detects MFA1A00 cable as a 100G-KR4 cable incorrectly. But the earlier generation of its NIC, Connect X-6/Connect X-4 detects it correctly as a 100G-SR4 cable. Raised an enterprise support case with NVIDIA and they replied mentioning that the mlx5 driver on FreeBSD is developed and maintained by the FreeBSD community and asked me to reach out here. I brought up Ubuntu 24.X on the same system and with the same firmware version the base PF driver detects the cable as a 100G-SR4 cable as expected. Upon a quick scan of the Linux mlx5 driver, the following snippet from the below patch helped identify the right connector type: https://github.com/torvalds/linux/commit/5b4793f817452e478442684e6bba85bddb5a9345 MLX5_GET(ptys_reg, out, connector_type); Cable spec is attached for reference. Request the experts in the community to work on the official mlx5 driver patch for FreeBSD. Regards, Nirmal -- You are receiving this mail because: You are the assignee for the bug.
[Bug 132798] ggatec(8): ggated/ggatec connection slowdown problem
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=132798 --- Comment #8 from o...@j.email.ne.jp --- Created attachment 258756 --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=258756&action=edit Adjusted to apply to 14.2-RELEASE I applied and addressed rejected patch. It compiles on 14.2-RELEASE and only briefly tested. -- You are receiving this mail because: You are the assignee for the bug.
[Bug 284815] RCTL memoryuse limit is not working
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=284815 --- Comment #3 from jsml4thwwbid6...@protonmail.com --- Hello, I think I may be misunderstanding how it's supposed to work. The processes I see using excessive memory are items like PHP, or PostgreSQL. Both seem to use some kind of anonymous memory, which does not appear to be restricted by the rctl memoryuse rule. I've tried the vmemoryuse option, and that causes resources to be restricted in a way that breaks normal operation. Also, the values it shows are pretty crazy. I've seen over a TB in memory usage reported, on a system with 128GB. The accounting seems really incorrect. Should I open a different ticket for this? Assuming I'm correct about the anonymous memory allocation being the issue, then is there a way to fully limit memory usage by user id? -- You are receiving this mail because: You are the assignee for the bug.
[Bug 283312] Kernel crash in sched_switch
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=283312 --- Comment #4 from Alexey Vyskubov --- I must mention that I turned off the swap encryption and have not had a single crash since then. -- You are receiving this mail because: You are the assignee for the bug.
[Bug 158230] libdisk(3) segfault in Open_Disk() with "ada0"
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=158230 Mark Linimon changed: What|Removed |Added Resolution|--- |Overcome By Events Assignee|b...@freebsd.org|bugmeis...@freebsd.org Status|Open|Closed --- Comment #3 from Mark Linimon --- ^Triage: I'm sorry that this PR did not get addressed in a timely fashion. By now, the version that it was created against is long out of support, and libdisk was removed 20121230. -- You are receiving this mail because: You are the assignee for the bug.
