Re: [PATCH 5.8 05/85] Revert "usbip: Implement a match function to fix usbip"

From: M. Vefa Bicakci
Date: Tue Oct 06 2020 - 10:14:40 EST


On 05/10/2020 18.26, Greg Kroah-Hartman wrote:
From: M. Vefa Bicakci <m.v.b@xxxxxxxxxx>

commit d6407613c1e2ef90213dee388aa16b6e1bd08cbc upstream.

This commit reverts commit 7a2f2974f265 ("usbip: Implement a match
function to fix usbip").

In summary, commit d5643d2249b2 ("USB: Fix device driver race")
inadvertently broke usbip functionality, which I resolved in an incorrect
manner by introducing a match function to usbip, usbip_match(), that
unconditionally returns true.

However, the usbip_match function, as is, causes usbip to take over
virtual devices used by syzkaller for USB fuzzing, which is a regression
reported by Andrey Konovalov.

Furthermore, in conjunction with the fix of another bug, handled by another
patch titled "usbcore/driver: Fix specific driver selection" in this patch
set, the usbip_match function causes unexpected USB subsystem behaviour
when the usbip_host driver is loaded. The unexpected behaviour can be
qualified as follows:
- If commit 41160802ab8e ("USB: Simplify USB ID table match") is included
in the kernel, then all USB devices are bound to the usbip_host
driver, which appears to the user as if all USB devices were
disconnected.
- If the same commit (41160802ab8e) is not in the kernel (as is the case
with v5.8.10) then all USB devices are re-probed and re-bound to their
original device drivers, which appears to the user as a disconnection
and re-connection of USB devices.

Please note that this commit will make usbip non-operational again,
until yet another patch in this patch set is merged, titled
"usbcore/driver: Accommodate usbip".

Cc: <stable@xxxxxxxxxxxxxxx> # 5.8: 41160802ab8e: USB: Simplify USB ID table match
Cc: <stable@xxxxxxxxxxxxxxx> # 5.8

Hello Greg,

Sorry for the lateness of this e-mail.

I had noted commit 41160802ab8e ("USB: Simplify USB ID table match") as a
prerequisite in the commit message, but I just realized that the commit
identifier refers to a commit in my local git tree, and not to the actual
commit in Linus Torvalds' git tree! I apologize for this mistake.

Here is the correct commit identifier:
0ed9498f9ecfde50c93f3f3dd64b4cd5414d9944 ("USB: Simplify USB ID table match")

Perhaps this is why the prerequisite commit was not cherry-picked to the 5.8.y
branch.

As a justification for the cherry-pick, commit 0ed9498f9ecf actually resolves
a bug. In summary, this commit works together with commit adb6e6ac20ee ("USB:
Also match device drivers using the ->match vfunc", which has been cherry-picked
as part of v5.8.6) and ensures that a USB driver's ->match function is also
called during the search for a more specialized/appropriate USB driver, in case
the driver in question does not have an id_table.

If I am to be the devil's advocate, however, then given that there is only one
specialized USB device driver ("apple-mfi-fastcharge"), which conveniently has
an id_table, and also given that usbip no longer has a match function, I also
realize that it may not be crucial to cherry-pick 0ed9498f9ecf as a prerequisite
commit.

I just wanted to bring this to your attention.

Thank you,

Vefa