Date: Mon, 17 Feb 2025 11:11:32 +0000 From: bugzilla-noreply@freebsd.org To: net@FreeBSD.org Subject: [Bug 265714] igc(4) drops link under high traffic Message-ID: <bug-265714-7501-J1RKMFNrR2@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-265714-7501@https.bugs.freebsd.org/bugzilla/> References: <bug-265714-7501@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D265714 --- Comment #20 from Phil Dennis-Jordan <phil@philjordan.eu> --- I upgraded to 14.2p1 a couple days ago and dealt with the dreaded drm-kmod = pkg problem by building the Intel graphics driver from ports. I've been hitting the NIC pretty hard with sustained loads since then which would previously have tripped it, and it's going strong so far. Relevant dm= esg boot output with the new version: igc0: <Intel(R) Ethernet Controller I226-V> mem 0x84200000-0x842fffff,0x84300000-0x84303fff at device 0.0 on pci2 igc0: EEPROM V2.17-0 eTrack 0x80000308 igc0: Using 1024 TX descriptors and 1024 RX descriptors igc0: Using 4 RX queues 4 TX queues igc0: Using MSI-X interrupts with 5 vectors igc0: Ethernet address: 48:21:0b:59:36:2d igc0: netmap queues/slots: TX 4/1024, RX 4/1024 Looking through the commit logs for igc(4), it seems interrupt management h= as changed between 14.1 and 14.2, with the introduction of "AIM". Does this AIM mode effectively work around a bug in the NIC's firmware or is there a bug = the non-AIM interrupt management in the driver that caused queues not to be processed in specific edge cases? (I didn't spot any other commits that mig= ht be relevant, but I'm new to this code; I also notice AIM remains disabled f= or speeds slower than gigabit, so could the issue still occur there?) --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-265714-7501-J1RKMFNrR2>