After building kernel 6.6.3, r8169 driver started acting funny First, it printed out error: [ 4.991530] r8169 0000:02:00.0 eth0: rtl_ep_ocp_read_cond == 0 (loop: 10, delay: 10000). Then my laptop just froze dead I managed to rebuild 6.6.3 image with r8169 as module, and tried to boot again. Thankfully, this time it did not crash the system, though internet still was not reachable with wired connection. Surprisingly, dhcpcd successfully ran and I got IP address, and LAN leds were active. It also printed out strange message 3 times: ipv6nd_sendrsprobe: No buffer space available And some strange things were happening with interface renaming (I would attach this as file) Reproducible: Always Steps to Reproduce: 1. Build kernel 6.6.3, having realtek ethernet controller managed by r8169 driver 2. Boot and try to establish internet connection 3. Actual Results: System froze with built in driver and had no internet connection with driver as module Expected Results: No freezing and properly working ethernet interface I also have to mention, that I have two ethernet interfaces on my laptop (Thinkpad P14s Gen 2)
Created attachment 875952 [details] dhcpcd and dmesg logs, addressing the problem
If your kernel has any of the CONFIG_RANDSTRUCT_ options set (other than NONE) try rebuilding & booting without.
(In reply to Holger Hoffstätte from comment #2) > If your kernel has any of the CONFIG_RANDSTRUCT_ options set (other than > NONE) try rebuilding & booting without. Thanks, I'll try. Any clue why it may interfere with eth driver?
See also: https://bugzilla.kernel.org/show_bug.cgi?id=218205
Can you test with gentoo-sources-6.6.3-r1, please?
(In reply to Mike Pagano from comment #5) > Can you test with gentoo-sources-6.6.3-r1, please? I was waiting for this revision, thanks, will try it out now
Just built and booted into 6.6.3-r1, seems to work fine. No failures for the moment If all runs good for day or two, I would close the bug as fixed