Comment 61 for bug 1931001

(In reply to Jarkko Nikula from comment #44)
> (In reply to stephane.poignant from comment #42)
> > I upgraded both from Debian Buster (kernel 4.19.194-3) to Bullseye
> > (5.10.46-5). No issue on the C3758, but i was faced with severe performance
> > regression on the C2758F.
> >
> Interesting, so was the 4.19 working on the C2758F without interrupt storm?

I haven't checked the /proc/interrupts when running 4.19 so i cannot tell for sure that the interrupts were not there. The performance regression was not there for sure. I can check this in a couple of weeks (server at a remote location with no oobm network).

Dmesg when running 4.19 shows it had interrupts enabled:

[ 0.000000] Linux version 4.19.0-17-amd64 (<email address hidden>) (gcc version 8.3.0 (Debian 8.3.0-6)) #1 SMP Debian 4.19.194-3 (2021-07-18)
[ 0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.19.0-17-amd64 root=/dev/mapper/vg1--hrbpsrv01-h--hrbpsrv01 ro quiet rd.luks.options=discard
...
[ 1.434097] Run /init as init process
[ 1.782787] dca service started, version 1.12.1
[ 1.783203] ismt_smbus 0000:00:13.0: enabling device (0000 -> 0002)
[ 1.796694] cryptd: max_cpu_qlen set to 1000
[ 1.801177] i801_smbus 0000:00:1f.3: enabling device (0000 -> 0003)
[ 1.801317] i801_smbus 0000:00:1f.3: SPD Write Disable is set
[ 1.801356] i801_smbus 0000:00:1f.3: SMBus using PCI interrupt
[ 1.805199] igb: Intel(R) Gigabit Ethernet Network Driver - version 5.4.0-k
[ 1.805202] igb: Copyright (c) 2007-2014 Intel Corporation.
[ 1.805246] igb 0000:00:14.0: enabling device (0000 -> 0002)
[ 1.816722] SSE version of gcm_enc/dec engaged.
...