linux-kernel - Re: ice driver not loading with 256 CPU's?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <68f4cde8-cb2f-f526-265d-2d5cdcfa53e2@praktifix.dwd.de>
Date:   Sun, 16 Oct 2022 10:08:51 +0000 (GMT)
From:   Holger Kiehl <Holger.Kiehl@....de>
To:     Jesse Brandeburg <jesse.brandeburg@...el.com>,
        Tony Nguyen <anthony.l.nguyen@...el.com>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        intel-wired-lan@...ts.osuosl.org, netdev@...r.kernel.org,
        linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: ice driver not loading with 256 CPU's?

On Sat, 15 Oct 2022, Holger Kiehl wrote:

> Hello,
>
> I have an AMD system with 2 sockets (each with a EPYC 7763 64-Core)
> with a total of 256 CPU's and a 4 port Intel 810 nic and get the
> following error during boot:
>
> Oct 15 10:53:35 hermes kernel: ice 0000:e2:00.1: The DDP package was 
> successfully loaded: ICE OS Default Package version 1.3.26.0
> Oct 15 10:53:35 hermes kernel: ice 0000:e2:00.1: not enough device MSI-X 
> vectors. requested = 260, available = 252
> Oct 15 10:53:35 hermes kernel: ice 0000:e2:00.1: ice_init_interrupt_scheme 
> failed: -34
> Oct 15 10:53:35 hermes kernel: ice: probe of 0000:e2:00.1 failed with error 
> -5
>
> Get this error when using default kernel from Alma9 or as above with
> kernel.org 6.0.2 kernel. Looking at the code
> (drivers/net/ethernet/intel/ice/ice_main.c ice_ena_msix_range() starting
> at line 3928) I would assume if I had less CPU's this would not be a problem.
>
Indeed, booting with nr_cpus=252 the driver loads:

Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: The DDP package was successfully loaded: ICE OS Default Package version 1.3.26.0
Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: PTP init successful
Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: DCB is enabled in the hardware, max number of TCs supported on this port are 8
Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: FW LLDP is disabled, DCBx/LLDP in SW mode.
Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: Commit DCB Configuration to the hardware
Oct 16 10:02:30 hermes kernel: ice 0000:e2:00.0: 252.048 Gb/s available PCIe bandwidth (16.0 GT/s PCIe x16 link)

Could you please fix the driver so it works with more then 252 CPU's?

Many thanks in advance.

Regards,
Holger