1 |
On Fri, 2020-03-27 at 19:07 -0400, Anthony G. Basile wrote: |
2 |
> On 3/27/20 3:17 PM, Alexis Ballier wrote: |
3 |
> > On Fri, 2020-03-27 at 08:03 -0400, Anthony G. Basile wrote: |
4 |
> > > On 3/26/20 9:25 PM, Joshua Kinard wrote: |
5 |
> > > > On 3/23/2020 04:21, Jaco Kroon wrote: |
6 |
> > > > > Hi, |
7 |
> > > > > |
8 |
> > > > > https://bugs.gentoo.org/713668 relates. |
9 |
> > > > > |
10 |
> > > > > * Searching for /usr/include/execinfo.h ... |
11 |
> > > > > sys-libs/glibc-2.29-r7 (/usr/include/execinfo.h) |
12 |
> > > > > |
13 |
> > > > > As I see I can either add an explicit depend on glibc which |
14 |
> > > > > I'd |
15 |
> > > > > prefer |
16 |
> > > > > not to. Or someone from the musl team could possibly assist |
17 |
> > > > > on |
18 |
> > > > > how to |
19 |
> > > > > get the backtrace() set of calls on musl please? |
20 |
> > > > > |
21 |
> > > > > Alternatively I need to add a test and simply path debug.c to |
22 |
> > > > > only |
23 |
> > > > > provide stub function for print_backtrace(FILE *fp) that just |
24 |
> > > > > does |
25 |
> > > > > fprintf(fp, "No backtrace() available to print a |
26 |
> > > > > backtrace.\n"); |
27 |
> > > > > |
28 |
> > > > > Suggestions? |
29 |
> > > > > |
30 |
> > > > > Kind Regards, |
31 |
> > > > > Jaco |
32 |
> > > > |
33 |
> > > > Some quick searching on google, it looks like the cleanest fix |
34 |
> > > > for |
35 |
> > > > that bug |
36 |
> > > > is dahdi-tools needs to be patched to only include execinfo.h |
37 |
> > > > or |
38 |
> > > > only use |
39 |
> > > > backtrace() on glibc-based systems, and that patch then sent to |
40 |
> > > > the |
41 |
> > > > dahdi-tools upstream developers for inclusion in a future |
42 |
> > > > release. That |
43 |
> > > > way, we're not dragging that patch around forever in the tree |
44 |
> > > > or in |
45 |
> > > > the musl |
46 |
> > > > overlay. |
47 |
> > > > |
48 |
> > > > It also doesn't look like musl itself will ever implement |
49 |
> > > > execinfo.h or |
50 |
> > > > backtrace(), per this message in 2015 from the lead musl |
51 |
> > > > developer: |
52 |
> > > > https://www.openwall.com/lists/musl/2015/04/09/3 |
53 |
> > > > |
54 |
> > > |
55 |
> > > Correct. I've been adding -standalone packages to provide for |
56 |
> > > features |
57 |
> > > like fts, obstack, argp,etc. which are bundled into glibc but not |
58 |
> > > really |
59 |
> > > under the POSIX standard. |
60 |
> > > |
61 |
> > > So either we patch packages to turn off backtrace() or we add |
62 |
> > > libunwind-standalone to the tree. |
63 |
> > > |
64 |
> > |
65 |
> > BTW, we had libexecinfo for fbsd, which seems also present in |
66 |
> > alpine: |
67 |
> > https://pkgs.alpinelinux.org/package/edge/main/x86/libexecinfo |
68 |
> > |
69 |
> > |
70 |
> |
71 |
> Had? Is it in the tree now or should I look into adding it? |
72 |
|
73 |
Restoring it: https://bugs.gentoo.org/683284 |