Bug 280972 - hms(4): Broken after 13d00a43cba4c35fcc9e0ab567baf530126a6348
Summary: hms(4): Broken after 13d00a43cba4c35fcc9e0ab567baf530126a6348
Status: Closed DUPLICATE of bug 280290
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: 15.0-CURRENT
Hardware: Any Any
: --- Affects Only Me
Assignee: Emmanuel Vadot
URL:
Keywords: regression
Depends on:
Blocks:
 
Reported: 2024-08-21 11:08 UTC by Mikael Urankar
Modified: 2024-10-27 17:36 UTC (History)
3 users (show)

See Also:

Attachments
bad (84.16 KB, text/plain)
2024-08-21 13:44 UTC, Mikael Urankar 		no flags Details
good (88.23 KB, text/plain)
2024-08-21 13:44 UTC, Mikael Urankar 		no flags Details
wait_for_reset.patch (8.04 KB, patch)
2024-09-15 11:20 UTC, Vladimir Kondratyev 		no flags Details | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Mikael Urankar freebsd_committer freebsd_triage 2024-08-21 11:08:28 UTC 
My touchpad doesn't work anymore after https://github.com/freebsd/freebsd-src/commit/13d00a43cba4c35fcc9e0ab567baf530126a6348
It works fine if I revert this commit.

This is a Dell Precision 3520

iichid0: <DLL07A9:00 044E:120B I2C HID device> at addr 0x2c irq 51 on iicbus7
hidbus0: <HID bus> on iichid0

hms0: <DLL07A9:00 044E:120B Mouse> on hidbus0
hms0: 3 buttons and [XYWH] coordinates ID=1

Not sure if it's the same problem as bug #280290, creating a new one.
Comment 1 Emmanuel Vadot freebsd_committer freebsd_triage 2024-08-21 13:25:50 UTC 
Can you share a verbose dmesg with and without this commit ?
Comment 2 Mikael Urankar freebsd_committer freebsd_triage 2024-08-21 13:44:40 UTC 
Created attachment 252985 [details]
bad
Comment 3 Mikael Urankar freebsd_committer freebsd_triage 2024-08-21 13:44:53 UTC 
Created attachment 252986 [details]
good
Comment 4 Mikael Urankar freebsd_committer freebsd_triage 2024-08-22 09:03:52 UTC 
CC'ing wulf

I've recompiled a kernel with IICHID_DEBUG and did what wulf@ asked in https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=280290#c2


without 13d00a43cba4 (working touchpad):
hkbd0: detached 
iichid0: iichid device close
iichid0: HID command I2C_HID_CMD_SET_POWER(1)
hkbd0: detached
hms0: detached
hidbus0: detached
iichid0: detached 
iicbus7: <unknown card> at addr 0x2c
iichid0: <DLL07A9:00 044E:120B I2C HID device> at addr 0x2c irq 51 on iicbus7
hidbus0: <HID bus> on iichid0
hms0: <DLL07A9:00 044E:120B Mouse> on hidbus0
hms0: 3 buttons and [XYWH] coordinates ID=1
hkbd0: <DLL07A9:00 044E:120B Keyboard> on hidbus0
kbd2 at hkbd0

after issuing kldunload iichid, the touchpad stopped working.




with 13d00a43cba4 (non working touchpad):
hkbd0: detached
iichid0: iichid device close
hms0: detached
hidbus0: detached
iichid0: detached
iicbus7: <unknown card> at addr 0x2c 
iichid0: <DLL07A9:00 044E:120B I2C HID device> at addr 0x2c irq 51 on iicbus7 
hidbus0: <HID bus> on iichid0
hms0: <DLL07A9:00 044E:120B Mouse> on hidbus0 
hms0: 3 buttons and [XYWH] coordinates ID=1 
hkbd0: <DLL07A9:00 044E:120B Keyboard> on hidbus0
kbd2 at hkbd0
Comment 5 Vladimir Kondratyev freebsd_committer freebsd_triage 2024-09-15 11:20:24 UTC 
Created attachment 253580 [details]
wait_for_reset.patch

I guess that compiling driver into kernel changes a winner of a race between HID hardware reset and iichid children initialization.

Test attached patch. Amongst other things it adds a waiting for HID device reset acknowledge that fixes the race.
Comment 6 Mikael Urankar freebsd_committer freebsd_triage 2024-09-15 17:05:21 UTC 
(In reply to Vladimir Kondratyev from comment #5)
It's still broken.
Comment 7 Mikael Urankar freebsd_committer freebsd_triage 2024-10-27 17:36:26 UTC 
*** This bug has been marked as a duplicate of bug 280290 ***