[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <CALAqxLW2R4d=Zm=TKbFprN-uYrerL1oCYsVC3VedEKtW0gCsyA@mail.gmail.com>
Date: Tue, 14 Apr 2020 23:44:04 -0700
From: John Stultz <john.stultz@...aro.org>
To: Felipe Balbi <balbi@...nel.org>
Cc: Josh Gao <jmgao@...gle.com>, YongQin Liu <yongqin.liu@...aro.org>,
Anurag Kumar Vulisha <anurag.kumar.vulisha@...inx.com>,
Yang Fei <fei.yang@...el.com>,
Thinh Nguyen <thinhn@...opsys.com>,
Tejas Joglekar <tejas.joglekar@...opsys.com>,
Andrzej Pietrasiewicz <andrzej.p@...labora.com>,
Jack Pham <jackp@...eaurora.org>, Todd Kjos <tkjos@...gle.com>,
Greg KH <gregkh@...uxfoundation.org>,
Linux USB List <linux-usb@...r.kernel.org>,
lkml <linux-kernel@...r.kernel.org>
Subject: More dwc3 gadget issues with adb
Hey Felipe,
Last week or so, a change[1] in AOSP to adbd seemingly uncovered
another issue with dwc3 gadget scatter-gather support on HiKey960.
Interestingly it doesn't seem to affect the Dragonboard 845c, which
uses the same dwc3 driver and has had its own issues in the past.
The behavior is the same as we saw last time around with both devices.
After booting the device, running "adb logcat -d" (or really any adb
command that transfers more than a trivial amount of data) on the host
will result in the adb output seeming to stall. Any further adb
invocations to the device will hang indefinitely.
I've captured trace events for before the change (works), after the
change (broken), and after the change with the sg_enabled flag turned
off (which works around the problem).
Let me know if there is anything else useful for me to share.
thanks
-john
[1] https://android.googlesource.com/platform/system/core/+/0871824de6c5b868f2a2e67c9b5673cb2181b4c6
Download attachment "hikey960.tar.xz" of type "application/octet-stream" (92332 bytes)
Powered by blists - more mailing lists