1 |
On Fri, Nov 11, 2005 at 11:33:25PM +0900, Jason Stubbs wrote: |
2 |
> On Friday 11 November 2005 22:53, Brian Harring wrote: |
3 |
> > Hola, |
4 |
> > |
5 |
> > Short version is that via bug 112082 plus glibc upstream doing |
6 |
> > something stupid, a lovely bug has reared it's head. Basically, |
7 |
> > users upgrading from glibc-2.3.5.200*.ebuild have libc.so-2.3.90.so, |
8 |
> > and glibc-2.3.6.ebuild has libc.so-2.3.6.so . |
9 |
> > |
10 |
> > ldconfig views 2.3.90 as greater then 2.3.6; during merge, portage |
11 |
> > views 2.3.5.200* -> 2.3.6 as an upgrade, and triggers an ldconfig |
12 |
> > call. Said call, and said upstream weird lib versioning results in |
13 |
> > ld.so.6 being reset from libc-2.3.6.so to libc-2.3.90.so , which |
14 |
> > obviously gets a bit screwed up when unmerge comes around and yanks |
15 |
> > the lib. |
16 |
> > |
17 |
> > Az split off a patch for it (after lots of fun digging) to correct it; |
18 |
> > we're going to need a rc8 covering this one offhand, since it's |
19 |
> > invalid linking in certain cases. |
20 |
> > |
21 |
> > Thoughts/complaints/issues/further testing? |
22 |
> |
23 |
> Short answer: No. |
24 |
Short response: reconsider. |
25 |
;) |
26 |
|
27 |
> Long answer: This is not a regression; you'll find the same problem in stable. |
28 |
> It's an area where a slight error can break lots of things that are even |
29 |
> slightly non-standard. This case where the bug is occurring is very |
30 |
> non-standard and has not cropped up before (or at least hasn't been brought |
31 |
> to anybody's attention) in the time since I (and you) have been with the |
32 |
> project. Lastly, 2.3.5.200* is/was hard masked. |
33 |
|
34 |
Yah it's a corner case, and yah few people are potential bitten by it |
35 |
right now- that said, the severity of being bitten by it is why this |
36 |
needs be released. Think about it, it's a cockup in our lib |
37 |
handling... an area that needs to run pretty much perfect. A corner |
38 |
case being spotted, even if upstream did something idiotic isn't |
39 |
particular acceptable in that chunk of code (imo). |
40 |
|
41 |
Regression or not, I'm after having a fix deployed as quickly as |
42 |
possible. Honestly, this reminds me of the early days of the daemon |
43 |
where occasionally it would lose the env and install nothing- didn't |
44 |
matter what the cause was, what mattered was that portage would go and |
45 |
rape my glibc installation, hosing my box. Potential for hosing the |
46 |
machine (correctable or not) via portage is a big no no from where |
47 |
I'm sitting. Bugs of this sort are more then capable of driving away |
48 |
a potential user/dev who has the misfortune of hitting it (masked or |
49 |
otherwise). |
50 |
|
51 |
> My preference would be to put the patch into trunk and release .54_pre1 within |
52 |
> the next 24 hours. |
53 |
|
54 |
Releasing .54 == releasing cache, pre/post patch. Yes getting those |
55 |
out for users to use is good, but see my earlier comments about |
56 |
sliding bug fixes in with large features. |
57 |
|
58 |
.53.1 strikes me as a better route; we're tagging for a reason, not too |
59 |
hard to pop back and make the change. |
60 |
~harring |