mirrors/git-annex
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

followup

Browse source
This commit is contained in:
Joey Hess 2018-11-12 14:54:50 -04:00
parent e4125318d8
commit 563f8b7c0f
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38
1 changed files with 23 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

23
doc/bugs/multiple_ssh_prompts__44___and_thread_blocked_indefinitely_in_an___63____63____63___transaction/comment_7_44d25c37e26e71b2aff759a0d39c5803._comment Normal file
View file

@ -0,0 +1,23 @@
[[!comment format=mdwn
username="joey"
subject="""comment 7"""
date="2018-11-12T17:54:07Z"
content="""
[2018-11-12 12:25:05.612916] read: ssh [\"-o\",\"BatchMode=true\",\"-S\",\".git/annex/ssh/yoh@falkor.datalad.org\",\"-o\",\"ControlMaster=auto\",\"-o\",\"ControlPersist=yes\",\"-n\",\"-T\",\"yoh@falkor.datalad.org\",\"true\"]
[2018-11-12 12:25:05.712832294] process done ExitFailure 255
git-annex: thread blocked indefinitely in an STM transaction
So sshing without letting it prompt for a password has failed.
The next thing git-annex should do at that point is take a lock
to prevent concurrent prompts (which does not use STM),
ask concurrent-output to hide any regions from the display
(which does use STM), and run ssh again to prompt.
It's apparently crashing before it can run ssh again. (Although I'm not really
sure if it is, it's also possible it overwrote some of the debug output while
trying to hide console regions. Please capture a complete typescript.)
It could be that concurrent-output's waitDisplayChange is buggy,
with some kind of race. But this does not seem to be behaving like a race condition,
does it? It happens every time?
"""]]