Issue metadata
Sign in to add a comment
|
Regression: Modem information and connection status are not displayed after running 'modem status' in crosh terminal when dongle is connected |
||||||||||||||||||||||
Issue descriptionChrome Version:69.0.3457.0/10783.0.0 dev channel Daisy,Quawks OS: Chrome OS What steps will reproduce the problem? (1)Login to the device. (2)Insert the pre-activated mobile dongle to the USB slot. (3)Open the network panel and wait for the mobile network to show up (4)Open a browser window and press Ctrl + alt + t (this will open a crosh terminal) (5)Run'modem status' and observe Expected:Modem status should show the modem information and the connection status. Actual:Nothing is displayed after running 'modem status' in command line This is a Regression issue seen from as same is working fine in Attached the screencast for reference..
,
Jul 13
Issue is still seen on Dev RC 69.0.3486.0/10866.1.0 and issue is working fine in latest M-68 beta
,
Jul 13
Issue is seen only in Verified mode. Unable to reproduce the issue in Dev mode.
,
Jul 16
Hi Ben, I can replicate this issue locally. Sometimes ModemManager does not start and is causing this issue. Please let me know if you need any logs. Issue also happens on M68 on quwaks randomly.
,
Jul 26
Issue needs an owner and understanding of impact to users for RBB.
,
Jul 26
We expect end users to use the Settings UI to obtain information about the modem. The "modem status" command in crosh is mainly for advanced users / developers. Lowering the priority as I don't think it's a P1 issue.
,
Jul 26
,
Aug 3
This bug has an owner, thus, it's been triaged. Changing status to "assigned".
,
Sep 4
Friendly ping to get an update as it is marked as RBB. Thanks
,
Sep 4
,
Sep 4
,
Sep 7
Able to reproduce the issue on latest beta-69.0.3497.87/10895.49.0 Thanks!
,
Sep 8
here's my theory. don't have a device locally to test, and i don't want to screw up the DUT i've locked too much. i think this is related to the debugd sandboxing work that went in for R69: https://chromium-review.googlesource.com/1053426 we bind mount /run, but not /var/run. this wasn't a problem because /var/run was a symlink to /run, but that changed for R66 to be a bind mount: https://chromium-review.googlesource.com/906889 while we transitioned most things to use /run directly, glib still goes through /var/run: gio/gdbusaddress.c: ret = g_strdup ("unix:path=/var/run/dbus/system_bus_socket"); backing up a little, crosh's 'modem' is simply a dbus call to debugd which in turn runs `modem` which is a shell script from the shill project. that shell script runs `mmcli` to get details about connected modems, mmcli queries the modemmanager daemon via dbus, mmcli is a modemmanager tool, and modemmanager uses glib for things. if i run mmcli through minijail to get an env similar to debugd/shill which omits /var/run mount, we see it fails: # /sbin/minijail0 --profile=minimalistic-mountns -b /run -b /var -v -u shill-scripts -g shill-scripts -- /usr/bin/mmcli -L error: couldn't get bus: Could not connect: Permission denied if i add a manual /var/run bind mount, it starts passing: # /sbin/minijail0 --profile=minimalistic-mountns -b /run -b /var -b /var/run -v -u shill-scripts -g shill-scripts -- /usr/bin/mmcli -L Found 1 modems: /org/freedesktop/ModemManager1/Modem/3 [Fibocom] MBIM [2CB7:0007] so i guess we have to hack up glib to use /run directly. lets see how that plays out in M71.
,
Sep 8
also, i'm not sure this needs to be blocking any real release. "crosh" is a developer tool, not a user interface.
,
Sep 8
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/60b496ea934d8c1a4cd88cc521631d75663bf96d commit 60b496ea934d8c1a4cd88cc521631d75663bf96d Author: Mike Frysinger <vapier@chromium.org> Date: Sat Sep 08 20:32:45 2018 glib: use /run for dbus sockets /var/run might not be available, so make sure we use /run directly. BUG= chromium:853158 TEST=precq passes Change-Id: Ia01960976019b4e9b44717d3a2d9b5f1bee17b5b Reviewed-on: https://chromium-review.googlesource.com/1214625 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Ben Chan <benchan@chromium.org> [add] https://crrev.com/60b496ea934d8c1a4cd88cc521631d75663bf96d/dev-libs/glib/glib-2.52.3-r2.ebuild [modify] https://crrev.com/60b496ea934d8c1a4cd88cc521631d75663bf96d/dev-libs/glib/glib-2.52.3-r1.ebuild
,
Sep 10
Dropping blocker per comment 14.
,
Sep 10
backporting to any version should be easy since glib hasn't changed in a while. i'll ask for M70 and leave it to TPMs to evaluate how much this actually matters to the release.
,
Sep 11
Your change meets the bar and is auto-approved for M70. Please go ahead and merge the CL to branch 3538 manually. Please contact milestone owner if you have questions. Owners: benmason@(Android), kariahda@(iOS), geohsu@(ChromeOS), abdulsyed@(Desktop) For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Sep 11
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/9ed460d94b55a766d8961c61cc936991845422c7 commit 9ed460d94b55a766d8961c61cc936991845422c7 Author: Mike Frysinger <vapier@chromium.org> Date: Tue Sep 11 17:51:55 2018 glib: use /run for dbus sockets /var/run might not be available, so make sure we use /run directly. BUG= chromium:853158 TEST=precq passes Change-Id: Ia01960976019b4e9b44717d3a2d9b5f1bee17b5b (cherry picked from commit 60b496ea934d8c1a4cd88cc521631d75663bf96d) Reviewed-on: https://chromium-review.googlesource.com/1220107 Reviewed-by: Mike Frysinger <vapier@chromium.org> Commit-Queue: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> [add] https://crrev.com/9ed460d94b55a766d8961c61cc936991845422c7/dev-libs/glib/glib-2.52.3-r2.ebuild [modify] https://crrev.com/9ed460d94b55a766d8961c61cc936991845422c7/dev-libs/glib/glib-2.52.3-r1.ebuild
,
Sep 11
going to close this out for M-70. if people really want this in M-69, feel free to re-open and comment, but i think we should leave it at this considering all the other churn M-69 has to deal with. |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by harpreet@chromium.org
, Jun 15 2018