From: Paul Menzel <pmenzel@molgen.mpg.de>
To: "Jörg Rödel" <joro@8bytes.org>,
"Suravee Suthikulpanit" <suravee.suthikulpanit@amd.com>,
"Bjorn Helgaas" <bhelgaas@google.com>
Cc: iommu@lists.linux-foundation.org,
Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
x86@kernel.org, LKML <linux-kernel@vger.kernel.org>,
linux-pci@vger.kernel.org
Subject: Re: How to reduce PCI initialization from 5 s (1.5 s adding them to IOMMU groups)s
Date: Fri, 5 Nov 2021 13:04:22 +0100 [thread overview]
Message-ID: <0f86c80b-aa6b-e1fd-5fda-0a4e4093250d@molgen.mpg.de> (raw)
In-Reply-To: <de6706b2-4ea5-ce68-6b72-02090b98630f@molgen.mpg.de>
Dear Linux folks,
Am 05.11.21 um 12:56 schrieb Paul Menzel:
> On a PowerEdge T440/021KCD, BIOS 2.11.2 04/22/2021, Linux 5.10.70 takes
> almost five seconds to initialize PCI. According to the timestamps, 1.5
> s are from assigning the PCI devices to the 142 IOMMU groups.
>
> ```
> $ lspci | wc -l
> 281
> $ dmesg
> […]
> [ 2.918411] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
> [ 2.933841] ACPI: Enabled 5 GPEs in block 00 to 7F
> [ 2.973739] ACPI: PCI Root Bridge [PC00] (domain 0000 [bus 00-16])
> [ 2.980398] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI HPX-Type3]
> [ 2.989457] acpi PNP0A08:00: _OSC: platform does not support [LTR]
> [ 2.995451] acpi PNP0A08:00: _OSC: OS now controls [PME PCIeCapability]
> [ 3.001394] acpi PNP0A08:00: FADT indicates ASPM is unsupported,
> using BIOS configuration
> [ 3.010511] PCI host bridge to bus 0000:00
> […]
> [ 6.233508] system 00:05: [io 0x1000-0x10fe] has been reserved
> [ 6.239420] system 00:05: Plug and Play ACPI device, IDs PNP0c02 (active)
> [ 6.239906] pnp: PnP ACPI: found 6 devices
> […]
> [ 6.989016] pci 0000:d7:05.0: disabled boot interrupts on device [8086:2034]
> [ 6.996063] PCI: CLS 0 bytes, default 64
> [ 7.000008] Trying to unpack rootfs image as initramfs...
> [ 7.065281] Freeing initrd memory: 5136K
> […]
> [ 7.079098] DMAR: dmar7: Using Queued invalidation
> [ 7.083983] pci 0000:00:00.0: Adding to iommu group 0
> […]
> [ 8.537808] pci 0000:d7:17.1: Adding to iommu group 141
> [ 8.571191] DMAR: Intel(R) Virtualization Technology for Directed I/O
> [ 8.577618] PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
> […]
> ```
>
> Is there anything that could be done to reduce the time?
I created an issue at the Kernel.org Bugzilla, and attached the output
of `dmesg` there [1].
Kind regards,
Paul
[1]: https://bugzilla.kernel.org/show_bug.cgi?id=214953
next prev parent reply other threads:[~2021-11-05 12:04 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-11-05 11:56 How to reduce PCI initialization from 5 s (1.5 s adding them to IOMMU groups) Paul Menzel
2021-11-05 12:04 ` Paul Menzel [this message]
2021-11-05 18:53 ` Bjorn Helgaas
2021-11-06 10:42 ` Paul Menzel
2021-11-09 15:31 ` Robin Murphy
2021-11-09 20:32 ` Paul Menzel
2021-11-08 17:18 ` Krzysztof Wilczyński
2021-11-09 20:25 ` Paul Menzel
2021-11-09 23:10 ` Krzysztof Wilczyński
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=0f86c80b-aa6b-e1fd-5fda-0a4e4093250d@molgen.mpg.de \
--to=pmenzel@molgen.mpg.de \
--cc=bhelgaas@google.com \
--cc=bp@alien8.de \
--cc=iommu@lists.linux-foundation.org \
--cc=joro@8bytes.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=suravee.suthikulpanit@amd.com \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).