1 |
Zac Medico posted on Fri, 31 May 2013 22:49:02 -0700 as excerpted: |
2 |
|
3 |
> On 05/31/2013 10:36 PM, Duncan wrote: |
4 |
>> As in subject, is portage bin/usr-bin merge safe? |
5 |
>> |
6 |
>> It appears most of my clashing files are /usr/bin/* -> /bin/* symlinks. |
7 |
>> (That's just bin, I've not looked at sbin.) |
8 |
> |
9 |
> I haven't tried it, but it should work just fine. Portage has always |
10 |
> supported directory symlinks like these. I haven't heard any recent |
11 |
> complaints regarding them. |
12 |
|
13 |
As the attribution says, I'm resurrecting a thread from 2013... |
14 |
|
15 |
I set up a merged /usr/bin -> /bin (and sbin -> bin, and /usr -> .) soon |
16 |
after that, with very few problems, usually ebuilds doing unconditional |
17 |
rms in postinst or the like, until recently... |
18 |
|
19 |
[I'll likely file this as a bug as well, but thought I'd post a followup |
20 |
to the old thread here, first. I'm kinda busy troubleshooting the |
21 |
unrelated bug that triggered the coreutils expression of this bug for me, |
22 |
ATM.] |
23 |
|
24 |
Something recently changed, as now I'm having many more problems, so far |
25 |
with four packages, glibc (!!), coreutils (!!), nano, and shadow, |
26 |
installing symlinks that ultimately point to themselves. The glibc one |
27 |
is of course critical as it breaks pretty much the entire system right |
28 |
away, the coreutils set is critical due to the number of frequently used |
29 |
binaries it breaks, and I'm lucky I discovered nano before needing it as |
30 |
a low-dep fallback editor. Being a single-user system I don't so often |
31 |
use passwd, but like nano, it's one of those things that when it's |
32 |
needed, it's REALLY needed. |
33 |
|
34 |
From my current installmask file: |
35 |
|
36 |
# 2017.1112 glibc: libm-2.*.so due to /usr -> . symlink, |
37 |
# symlink overwrites the lib it points to! |
38 |
INSTALL_MASK=" |
39 |
$INSTALL_MASK |
40 |
/usr/lib64/libm-2.*.so |
41 |
" |
42 |
|
43 |
# 2017.1207 coreutils symlinks that overwrite their binaries |
44 |
INSTALL_MASK=" |
45 |
$INSTALL_MASK |
46 |
/usr/bin/basename |
47 |
/usr/bin/chroot |
48 |
/usr/bin/cut |
49 |
/usr/bin/dir |
50 |
/usr/bin/dirname |
51 |
/usr/bin/du |
52 |
/usr/bin/env |
53 |
/usr/bin/expr |
54 |
/usr/bin/head |
55 |
/usr/bin/mkfifo |
56 |
/usr/bin/mktemp |
57 |
/usr/bin/readlink |
58 |
/usr/bin/seq |
59 |
/usr/bin/sleep |
60 |
/usr/bin/sort |
61 |
/usr/bin/tail |
62 |
/usr/bin/touch |
63 |
/usr/bin/tr |
64 |
/usr/bin/tty |
65 |
/usr/bin/uname |
66 |
/usr/bin/vdir |
67 |
/usr/bin/wc |
68 |
/usr/bin/yes |
69 |
" |
70 |
# 2017.1207 shadow, nano symlinks |
71 |
INSTALL_MASK=" |
72 |
$INSTALL_MASK |
73 |
/usr/bin/nano |
74 |
/usr/bin/passwd |
75 |
" |
76 |
|
77 |
So what changed in portage that previously prevented the /usr/* symlinks |
78 |
from overwriting the non-usr binaries, but now allows the overwrites to |
79 |
go ahead, breaking the system? |
80 |
|
81 |
Note that I ran into the glibc library symlink issue first. I ran into |
82 |
the coreutils issue after a bad upgrade (unrelated, I think) broke X, |
83 |
forcing me back to a backup and I started upgrading a few packages at a |
84 |
time from binpkg, to see which one broke X again. When I got to |
85 |
coreutils, the qmerge phase broke half way thru as a binary was replaced |
86 |
by a symlink to itself. I'm not sure why it triggered in binpkg install |
87 |
but not when I had originally installed it on the system, but it may be |
88 |
due to the fact that I normally run parallel merges so the system is |
89 |
heavily loaded during normal merge, while with the binpkg merge it |
90 |
wasn't, thus implying a race condition of some sort. I discovered the |
91 |
nano and shadow/passwd issues, seeing their binaries were broken symlinks |
92 |
to themselves, when fixing the coreutils issue. I've no idea when they |
93 |
happened. |
94 |
|
95 |
-- |
96 |
Duncan - List replies preferred. No HTML msgs. |
97 |
"Every nonfree program has a lord, a master -- |
98 |
and if you use the program, he is your master." Richard Stallman |