Closed txtsd closed 3 years ago
If you want to, you can narrow this down to the offending commit via git bisect
.
git clone https://github.com/brndnmtthws/conky
git bisect
stuffs (Look this up). It's something like this.
git bisect start
git checkout v1.11.4 && git bisect bad
git checkout v1.11.3 && git bisect good
Here we go!
λ git bisect bad
c352069b31b2dbb4d87503603d6c6d3ddd517df6 is the first bad commit
commit c352069b31b2dbb4d87503603d6c6d3ddd517df6
Author: Kushagra Sinha <sinhak@google.com>
Date: Sun Mar 3 17:45:02 2019 +0100
Add support for nested |exec| expressions
Nested exec expressions like ${execp echo '${execp echo hi}'} are not
evaluated correctly because parsing expressions like these generates and
registers new callbacks but we never wait for these callbacks to
complete before returning the result of the evaluation.
Fix this by re-evaluating expressions after running all pending
callbacks.
src/conky.cc | 11 +++++++++
tests/CMakeLists.txt | 2 +-
tests/test-conky.cc | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 75 insertions(+), 1 deletion(-)
create mode 100644 tests/test-conky.cc
Any updates on this? I still haven't been able to use anything after 1.11.3
@txtsd I'll try and look into it within the next week or two.
Thanks
Is it fixed ¿
Not yet, unfortunately :cry:
It seems that I've encountered this one with a current build of v1.11.6 on gentoo. Starting conky with --daemonize or passing in background=true reliably "hides", or better "shrinks" the conky window to a tiny little square... terminating is only possible using kill -KILL.
conky -p5 -m0 --daemonize:
conky -p5 -m0:
@brndnmtthws Any headway into this? I'd like to update at some point.
@txtsd Tried your .conkyrc. No display until I change to background=false.
@txtsd Tried your .conkyrc. No display until I change to background=false.
No point in using conky unless it is true. The offending commit has already been found though, and I'd like for an actual resolution.
I believe #1093 will resolve this. If someone could test and confirm for me, that would be excellent.
It looks like the change that caused the regression wasn't quite right, but it seems to be working as intended now. The call to conky::run_all_callbacks()
was blocking the main thread, resulting in the appearance of Conky to stop. This only happened with nested expressions, like ${if_...}
, as near as I could tell.
Compiled the branch in that PR and it works! Will you release a new build with this bugfix?
Issue
conky
does not display anything on my screen even though it runs and forks to background, albeit slower than with1.11.3
Information
Backtrace.
There's no crash so the
gdb
looks normal.Debug flag output
Here is the output of
conky -DD
Desktop with 1.11.3
Desktop with 1.11.4