On Dec 31, 2010, at 2:54 AM, Cyril Hrubis wrote:

> Hi!
>>      *facepalm* git and I really don't get along. I tried to merge the 
>> change from ltp-broom, that pulled in a bunch of stuff I didn't want, reset 
>> via reset --hard, then checked out the one file from ltp-broom, which 
>> worked, and when I pushed everything -- it pushed all of my changes from 
>> ltp-broom back to master. So I suppose the question now is to continue with 
>> things as-is, or revert the commit. I'm guessing folks will want the latter, 
>> but I'm unfamiliar enough with merging and reverting with git because it's 
>> user friendliness that I'm concerned about doing more harm than good trying 
>> to fix things.
> 
> Just now we need to get git to some sane state so that patches created
> against latest git could be merged later. But I'm not sure if reseting
> back is better than fixing the problems created by merge. If all the
> problems are created by merging files that were changed in both master
> and ltp-broom then IMHO fixing these would be better way to go.
> 
> FYI: just now compilation dies with:
> 
> gcc   -L../../../../lib  test_arch.o   -o test_arch_debug
> test_arch.o: In function `main':
> /home/metan/Work/git/ltp-dev/testcases/commands/ade/objdump/test_arch.c:17:
> undefined reference to `tst_exit'
> collect2: ld returned 1 exit status
> make[4]: *** [test_arch_debug] Error 1
> make[4]: Leaving directory
> `/home/metan/Work/git/ltp-dev/testcases/commands/ade/objdump'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
> `/home/metan/Work/git/ltp-dev/testcases/commands/ade'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory
> `/home/metan/Work/git/ltp-dev/testcases/commands'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/home/metan/Work/git/ltp-dev/testcases'
> make: *** [testcases-all] Error 2

This was fixed, but there's a lot more to go.

>>      I need to go back home (where I don't have internet via my laptop right 
>> now and won't for the next couple of days), but I'll continue to work on 
>> verifying everything and fix regressions if and when I find them, and I'll 
>> try to pop over to work whenever possible to push code back to the git repo.
> 
> I hope that lack of the internet connection in your home is only
> temporary, as so far you did the commits and there is a already some
> backlog of waiting patches. I know that it's holidays now but I'm
> allready starting to lose track of the issues waiting for commit.

Yeah, it was. Moving to a new place after coming back from a trip is a pain in 
the ass.

> If you need some help with that, don't hesitate to ask.

Mmk. There are 400 or so files that need fixing under syscalls (if you take out 
the set -e in generic_trunk_target.inc and run make -k with -Werror and -Wall, 
you'll see all of the issues). I turned on -Werror by default in config.mk.in 
so issues should crop up a bit easier now than before -- the only problem is 
with areas I didn't touched in the tree over the past 2 weeks.

There are a handful of other spots where my search and replace scripts screwed 
up too (leftover braces after parse_msg_options, etc). Most of this stuff is 
easy to identify -- it's just a matter of working through all of the items one 
by one (which won't be hard -- just time consuming... arg).

I apologize for the breakage, but it should be good to go in the next day or so 
for broader testing.

Thanks,
-Garrett
------------------------------------------------------------------------------
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and, 
should the need arise, upgrade to a full multi-node Oracle RAC database 
without downtime or disruption
http://p.sf.net/sfu/oracle-sfdevnl
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to