FS#7083 - dependencies not removed after md5sum check fail

Attached to Project: Pacman
Opened by Roman Kyrylych (Romashka) - Sunday, 06 May 2007, 15:52 GMT
Last edited by Roman Kyrylych (Romashka) - Tuesday, 05 June 2007, 14:29 GMT
Task Type Bug Report
Category makepkg
Status Closed
Assigned To Aaron Griffin (phrakture)
Roman Kyrylych (Romashka)
Dan McGee (toofishes)
Architecture All
Severity Low
Priority Normal
Reported Version 3.0.3
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

[roman@home nautilus-share]$ makepkg -Ssrf
==> Entering fakeroot environment
==> Making package: nautilus-share 0.7.1-1 (Ндл Тра 6 18:50:47 EEST 2007)
==> Checking Runtime Dependencies...
==> Checking Buildtime Dependencies...

==> Installing missing dependencies...
ERROR: ld.so: object 'libfakeroot.so' from LD_PRELOAD cannot be preloaded: ignored.
resolving dependencies... done.
looking for inter-conflicts... done.

Targets: intltool-0.35.5-1

Total Package Size: 0,10 MB

Proceed with installation? [Y/n]
checking package integrity... done.
cleaning up... done.
(1/1) checking for file conflicts [#####################] 100%
(1/1) installing intltool [#####################] 100%
==> Retrieving Sources...
-> Using cached copy of nautilus-share-0.7.1.tar.gz
-> Found usershare.conf in build dir
==> Validating source files with md5sums
nautilus-share-0.7.1.tar.gz ... Passed
usershare.conf ... FAILED
==> ERROR: One or more files did not pass the validity check!
[roman@home nautilus-share]$ sudo pacman -R intltool
checking dependencies... done.
(1/1) removing intltool [#####################] 100%
[roman@home nautilus-share]$
This task depends upon

Closed by  Roman Kyrylych (Romashka)
Tuesday, 05 June 2007, 14:29 GMT
Reason for closing:  Fixed
Additional comments about closing:  in git
Comment by Dan McGee (toofishes) - Sunday, 06 May 2007, 23:01 GMT
This whole "makepkg is doing way too much" bug is rearing its ugly head again. :P

I'm still not a big fan of makepkg doing all this. There are likely a lot of places that need checking for this stuff, so a patch would be much appreciated. Basically anywhere that "exit" is called in the script, we may need to run some cleanup stuff first.
Comment by Aaron Griffin (phrakture) - Friday, 11 May 2007, 17:31 GMT
We could also take advantage of bash's signal traps here, to basically register a C style 'atexit' function... might be too much
Comment by Dan McGee (toofishes) - Friday, 11 May 2007, 18:20 GMT
Once I get a decent amount of time those fakeroot only-when-required patches should get applied, and then we should work on these kind of cases.

I think an atexit wouldn't be too extreme, but it needs further thought.
Comment by Andrew Fyfe (space-m0nkey) - Wednesday, 30 May 2007, 18:27 GMT
I'll have I look into this and see if I can knock up a few ideas/patches.

Loading...