|
Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes. Set by lizmat on 24 May 2021. |
|||
|
00:02
reportable6 left
00:03
reportable6 joined
00:58
Kaiepi left,
Kaipi joined
01:05
Kaiepi joined,
Kaipi left
01:36
Kaiepi left
01:39
Kaiepi joined
01:59
squashable6 left
02:00
squashable6 joined
02:03
frost joined
03:58
evalable6 left,
linkable6 left
03:59
evalable6 joined
04:01
linkable6 joined
06:03
reportable6 left
06:04
reportable6 joined
|
|||
| Geth | MoarVM: a4d5949202 | (Stefan Seifert)++ | src/spesh/stats.c Fix memory corruption by accessing freed spesh stats. When cleaning up spesh stats that didn't get updated for a while, we free the MVMSpeshStats structure and remove the reference from the MVMStaticFrameSpesh object holding it. It may still be referenced from a MVMSpeshSimStackFrame however and when we continue processing this sim stack, we access some random memory. Look through all tc's spesh_sim_stacks for frames still referencing the MVMSpeshStats and just not free it if we find one. |
06:16 | |
| MoarVM: 0d3fc79e59 | niner++ (committed using GitHub Web editor) | src/spesh/stats.c Merge pull request #1511 from MoarVM/fix_spesh_stats_corruption Fix memory corruption by accessing freed spesh stats. |
|||
| MoarVM: 21f8caddb2 | (Stefan Seifert)++ | src/6model/serialization.c Fix "Collectable in a gen2 freelist accessed" when deserializing a method cache We deserialize the method cache hash itself. Then we run the work_loop to deserialize its entries. This may trigger GC. Since the cache at that point is not referenced by any root, we free it. Serialization and deserialization runs with allocation in gen2 enabled, so we ... (6 more lines) |
06:17 | ||
| MoarVM: d7a07bf321 | (Stefan Seifert)++ | src/6model/serialization.c Fix possible acces to a freed SC during deserialization While deserialized objects get allocated in gen2 directly, the same is not necessarily true for serialization contexts themselves. And having the allocate in gen2 directly flag active does not prevent GC runs in all cases (a different thread may allocate and trigger the GC). So when running the deserialization work loop, we need to take care to root the sc pointers that we continue to use later on. |
|||
| MoarVM: 65ed1ff246 | niner++ (committed using GitHub Web editor) | src/6model/serialization.c Merge pull request #1512 from MoarVM/fix_serialization_gc_issues Fix serialization gc issues |
|||
| MoarVM: ab2a01001f | (Tom Browder)++ | src/profiler/heapsnapshot.c Simplify preprocessor code |
06:19 | ||
| MoarVM: f8d5f30021 | (Tom Browder)++ | src/profiler/heapsnapshot.c Note only the gcc compiler is the one generating the warning |
|||
| MoarVM: 142f91176e | niner++ (committed using GitHub Web editor) | src/profiler/heapsnapshot.c Merge pull request #1503 from tbrowder/fix-heapsnapshot-warnings Simplify pre-processor code |
|||
|
07:29
patrickb joined
07:32
patrickb left
07:33
patrickb joined
08:55
evalable6 left,
linkable6 left
08:57
linkable6 joined
08:58
evalable6 joined
09:44
Kaiepi left
09:54
Kaiepi joined
|
|||
| Geth | MoarVM: 4fea2f9af8 | (Samantha McVey)++ | src/strings/shiftjis.c Fix ShiftJIS decoder bug: \r + error + replacement If the following are true decoding when ShiftJIS: * The previous character \r * We encounter a decoding error * We are using a replacement Then we encounter a case where things are wrong Input: "\r" + ERROR + "abc"; with replacement="hello" Expected: "\rhelloabc" Actual: "\rellohelloabc" |
10:22 | |
| MoarVM: 74e828f0a8 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/strings/shiftjis.c Merge pull request #1502 from samcv/fix-shiftjis-cr-bug Fix ShiftJIS decoder bug: \r + error + replacement |
|||
| lizmat | when I try to bump NQP to the latest version of Moar | 10:29 | |
| the build fails with: src/vm/moar/runner/main.c:4:10: fatal error: 'uv.h' file not found | |||
| since nothing was changed in nqp for the past 3 weeks or so, I assume this is something in Moar messing up ? | 10:30 | ||
|
10:54
linkable6 left,
evalable6 left
10:55
evalable6 joined
10:56
linkable6 joined
11:00
patrickb left
|
|||
| tbrowder | lizmat: check my PR just merged above #1503 | 11:12 | |
| it was in MoarVM | 11:13 | ||
| lizmat | weird... I've tried all the intermediate commits one by one, and after that, the final one also built ? | 11:33 | |
|
11:48
squashable6 left
11:49
squashable6 joined
12:02
reportable6 left
12:04
reportable6 joined
12:11
patrickb joined
|
|||
| tbrowder | i think my pr may have been built and tested without samcv's | 12:12 | |
| lizmat | well, they appear all good in the end | 12:13 | |
| tbrowder | gut! | ||
| lizmat | maybe it was a local issue with branch switching or so | ||
| running spectest now | |||
| patrickb | lizmat: I suspect github.com/MoarVM/MoarVM/commit/42...cb6ab9cdbd | ||
| lizmat | could | 12:14 | |
| be | |||
|
13:32
linkable6 left
13:33
linkable6 joined
14:14
squashable6 left
14:15
frost left
14:17
squashable6 joined
14:52
linkable6 left
14:55
linkable6 joined
15:00
linkable6 left
15:01
linkable6 joined
15:13
linkable6 left
15:15
linkable6 joined
15:23
linkable6 left
15:25
linkable6 joined
15:31
patrickb left
15:43
brrt joined
15:59
linkable6 left
16:00
linkable6 joined
17:45
brrt left
18:02
reportable6 left
18:04
reportable6 joined
18:36
discord-raku-bot joined
20:36
patrickb joined
21:43
patrickb left
|
|||