| 8 Feb 2024 |
 @aloisw:kde.org | There already is https://github.com/NixOS/nixpkgs/pull/216341 btw | 18:04:02 |
 raitobezarius | yeah but it's not doing the change in cc-wrapper | 18:04:56 |
| @aloisw:kde.org | It should probably also be discussed how useful frame pointers are when the debug symbols have already been lost, as is seemingly very common. | 18:05:23 |
| raitobezarius | well i wouldn't make frame pointers dependent on debug symbols | 18:09:43 |
| raitobezarius | we can make frame pointers now | 18:09:49 |
| raitobezarius | and still work towards dsyms no? | 18:09:53 |
| @aloisw:kde.org | I wasn't saying frame pointers are useless without debug info, you can probably still get some results from the symbol table alone. | 18:12:19 |
| raitobezarius | right I'm not saying you said that | 18:12:30 |
| raitobezarius | I'm not sure the discussion about usefulness without dsyms is important to make the change | 18:12:47 |
| @aloisw:kde.org | But frame pointers have a cost, and their benefit is at least somewhat reduced without debug info. | 18:12:54 |
| raitobezarius | right | 18:15:39 |
| raitobezarius | this should be benchmarked then | 18:15:45 |
| @aloisw:kde.org | I now remember this does not actually seem to be the case. The profiler frontend I use (Hotspot) is unable to augment the frame pointer data with debug info data. | 18:30:46 |
| @aloisw:kde.org | One of the reasons I use DWARF-based unwinding. | 18:31:10 |
| @aloisw:kde.org | But take my statements about profiling with a grain of salt in general, my use case is very specific (single-process, deterministic, CPU (or swap) bound). | 18:32:33 |
 @trofi:matrix.org | Default strip hook removes dwarf data, but not the symbols themselves. nixpkgs packages have almost all the symbols in all binaries and libraries. | 19:57:43 |
| @trofi:matrix.org | That 's for example why perf top reports all it's own symbols when it dies (instead of reporting just addresssssssses) :)
$ perf top -a
perf: Segmentation fault
-------- backtrace --------
/run/current-system/sw/bin/perf[0x626d62]
/nix/store/7jiqcrg061xi5clniy7z5pvkc4jiaqav-glibc-2.38-27/lib/libc.so.6(+0x3deb0)[0x7fe05da54eb0]
/run/current-system/sw/bin/perf(__dsos__findnew_link_by_longname_id+0x34b)[0x53633b]
/run/current-system/sw/bin/perf(map__new+0x35f)[0x55867f]
/run/current-system/sw/bin/perf(machine__process_mmap2_event+0xb7)[0x557957]
/run/current-system/sw/bin/perf(perf_tool__process_synth_event+0x7e)[0x59e3ee]
/run/current-system/sw/bin/perf(perf_event__synthesize_mmap_events+0x369)[0x59ec59]
/run/current-system/sw/bin/perf[0x59f8dc]
/run/current-system/sw/bin/perf[0x59fb4c]
/run/current-system/sw/bin/perf[0x59fc13]
/nix/store/7jiqcrg061xi5clniy7z5pvkc4jiaqav-glibc-2.38-27/lib/libc.so.6(+0x8b333)[0x7fe05daa2333]
/nix/store/7jiqcrg061xi5clniy7z5pvkc4jiaqav-glibc-2.38-27/lib/libc.so.6(+0x10defc)[0x7fe05db24efc]
| 19:59:59 |
| @trofi:matrix.org | (Oh, and perf crash is a gcc bug, I completely forgot that the fix was not yet backported to gcc-13 branch: https://gcc.gnu.org/PR111009) | 20:19:31 |
| @trofi:matrix.org | * (Oh, and perf crash is a gcc bug, I completely forgot that the fix was not yet backported to gcc-13 branch: https://gcc.gnu.org/PR111009; Proposed the workaround as https://github.com/NixOS/nixpkgs/pull/287310 ) | 20:54:49 |
| @trofi:matrix.org | binutils: 2.40 -> 2.41 update is ready for review (not sure if we need a hydra run against the branch before the merge): https://github.com/NixOS/nixpkgs/pull/283732 | 22:57:28 |
| 9 Feb 2024 |
 Philip Taron (UTC-8) | In reply to @trofi:matrix.org
binutils: 2.40 -> 2.41 update is ready for review (not sure if we need a hydra run against the branch before the merge): https://github.com/NixOS/nixpkgs/pull/283732 What sort of review would you like, trofi ? I looked through the set of changes and they do what they say on the tin. | 17:24:23 |
| @trofi:matrix.org | In reply to @philiptaron:matrix.org
What sort of review would you like, trofi ? I looked through the set of changes and they do what they say on the tin. I wonder if we should request a one-off hydra jobset or people confident it's already good enough and we should push it to staging ~as is. | 20:44:13 |
| Philip Taron (UTC-8) | I'd send it. | 20:48:55 |
| Philip Taron (UTC-8) | Gave it the old ✅ for whatever that's worth. | 21:01:15 |
