On 06/11/2020 10:36, Saravana Kannan wrote:
On Thu, Nov 5, 2020 at 9:09 PM Laurent Pinchart
<laurent.pinchart@xxxxxxxxxxxxxxxx> wrote:
Hi Saravana,
Thank you for working on this !
On Wed, Nov 04, 2020 at 03:23:37PM -0800, Saravana Kannan wrote:
The current implementation of fw_devlink is very inefficient because it
tries to get away without creating fwnode links in the name of saving
memory usage. Past attempts to optimize runtime at the cost of memory
usage were blocked with request for data showing that the optimization
made significant improvement for real world scenarios.
We have those scenarios now. There have been several reports of boot
time increase in the order of seconds in this thread [1]. Several OEMs
and SoC manufacturers have also privately reported significant
(350-400ms) increase in boot time due to all the parsing done by
fw_devlink.
So this patch series refactors fw_devlink to be more efficient. The key
difference now is the addition of support for fwnode links -- just a few
simple APIs. This also allows most of the code to be moved out of
firmware specific (DT mostly) code into driver core.
This brings the following benefits:
- Instead of parsing the device tree multiple times (complexity was
close to O(N^3) where N in the number of properties) during bootup,
fw_devlink parses each fwnode node/property only once and creates
fwnode links. The rest of the fw_devlink code then just looks at these
fwnode links to do rest of the work.
- Makes it much easier to debug probe issue due to fw_devlink in the
future. fw_devlink=on blocks the probing of devices if they depend on
a device that hasn't been added yet. With this refactor, it'll be very
easy to tell what that device is because we now have a reference to
the fwnode of the device.
- Much easier to add fw_devlink support to ACPI and other firmware
types. A refactor to move the common bits from DT specific code to
driver core was in my TODO list as a prerequisite to adding ACPI
support to fw_devlink. This series gets that done.
Tomi/Laurent/Grygorii,
If you can test this series, that'd be great!
I gave it a try, rebasing my branch from v5.9 to v5.10-rc2 first. On
v5.10-rc2 the kernel dies when booting due to a deadlock (reported by
lockdep, so hopefully not too hard to debug). *sigh*. Fortunately, it
dies after the fw_devlink initialization, so I can still report results.
Phew! For a sec I thought you said fw_devlink was causing a deadlock.
Before your series:
[ 0.743065] cpuidle: using governor menu
[ 13.350259] No ATAGs?
With your series applied:
[ 0.722670] cpuidle: using governor menu
[ 1.135859] No ATAGs?
That's a very clear improvement :-)
Thanks for testing. Great to hear it's helping!
Tested-by: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
I'll add it to my v2 series.
I've tried your series on top of
521b619acdc8 Merge tag 'linux-kselftest-kunit-fixes-5.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest
on am571x-idk
Before:
[ 0.049395] cpuidle: using governor menu
[ 1.654766] audit: type=2000 audit(0.040:1): state=initialized audit_enabled=0 res=1
[ 2.315266] No ATAGs?
[ 2.315317] hw-breakpoint: found 5 (+1 reserved) breakpoint and 4 watchpoint registers.
[ 2.315327] hw-breakpoint: maximum watchpoint size is 8 bytes.
...
[ 6.549595] EXT4-fs (mmcblk0p2): mounted filesystem with ordered data mode. Opts: (null)
[ 6.557794] VFS: Mounted root (ext4 filesystem) on device 179:26.
[ 6.574103] devtmpfs: mounted
[ 6.577749] Freeing unused kernel memory: 1024K
[ 6.582433] Run /sbin/init as init process
after:
[ 0.049223] cpuidle: using governor menu
[ 0.095893] audit: type=2000 audit(0.040:1): state=initialized audit_enabled=0 res=1
[ 0.102958] No ATAGs?
[ 0.103010] hw-breakpoint: found 5 (+1 reserved) breakpoint and 4 watchpoint registers.
[ 0.103020] hw-breakpoint: maximum watchpoint size is 8 bytes.
...
[ 3.518623] EXT4-fs (mmcblk0p2): mounted filesystem with ordered data mode. Opts: (null)
[ 3.526822] VFS: Mounted root (ext4 filesystem) on device 179:26.
[ 3.543128] devtmpfs: mounted
[ 3.546781] Freeing unused kernel memory: 1024K
[ 3.551463] Run /sbin/init as init process
So, it's much better. Thank you.
Tested-by: Grygorii Strashko <grygorii.strashko@xxxxxx>
--
Best regards,
grygorii
