{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":741636496,"defaultBranch":"main","name":"chapel","ownerLogin":"amritpan","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2024-01-10T19:56:33.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/32915938?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1704916606.31692","currentOid":""},"activityList":{"items":[{"before":"944920df90c0a88ba2fef12ba3154897fa740f7d","after":"0f759f827b17ad07fef523da7f93c213c63eaafd","ref":"refs/heads/main","pushedAt":"2024-03-18T21:22:59.000Z","pushType":"push","commitsCount":1082,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"CHANGES.md for Chapel 2.0 (#24531)\n\n[reviewed by @mppf and others piecemeal along the way]\r\n\r\nUpdate the changes list for the 2.0 release and fix some formatting\r\ninconsistencies in past entries.","shortMessageHtmlLink":"CHANGES.md for Chapel 2.0 (chapel-lang#24531)"}},{"before":"b92aac723b6f4343d00b8840df93acd0562e9508","after":"944920df90c0a88ba2fef12ba3154897fa740f7d","ref":"refs/heads/main","pushedAt":"2024-02-14T20:01:29.000Z","pushType":"push","commitsCount":119,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"Update normalized representation of domain queries to avoid array copies (#24416)\n\nIn https://github.com/chapel-lang/chapel/issues/24330 we observed that\r\nextra copies were occurring for initializers\r\nwritten to take an array by in-intent:\r\n\r\n```chpl\r\nproc ArrayWrapper.init(in a: [?d] ?t) {\r\n this.etype = t;\r\n this.rank = d.rank;\r\n this.shape = d.shape;\r\n this.a = a; // extra/unexpected copy here!\r\n\r\n writeln(d); // <- commenting this line out removes the extra copy\r\n}\r\n```\r\n\r\nThis was surprising since the 'in' intent rules are fairly well-defined\r\nand implemented.\r\n\r\nThe cause was that the normalize pass was transforming the writeln call\r\ninto something like ``writeln(a.domain)``. The late mention of 'a' would\r\nthen prevent the split-init logic from eliding the copy.\r\n\r\nThis PR changes the normalized representation of domain queries to\r\njust use a single 'const ref' temp that we set up at the beginning of\r\nthe function.\r\n\r\n[reviewed-by @mppf]","shortMessageHtmlLink":"Update normalized representation of domain queries to avoid array cop…"}},{"before":"9cd20851fd1248691beceb679e1b56a592e4f5c4","after":"b92aac723b6f4343d00b8840df93acd0562e9508","ref":"refs/heads/main","pushedAt":"2024-02-13T01:55:50.000Z","pushType":"push","commitsCount":134,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"GPU Privatization: handle a case where array record is LICM'd (#24370)\n\nhttps://github.com/chapel-lang/chapel/pull/24137 run into a problem with\r\n`--gpu-specialization`. With that mode, the loop that we end up\r\nlaunching as a kernel was subjected to aggressive LICM before we can\r\nfind pids in them. Because the PR initially tried to find array/domain\r\nrecords used in the loop. This PR expands the logic to also find indices\r\nused while indexing `chpl_privateObjects` array. The end result is that\r\nwe potentially capture some pids twice redundantly which may end up with\r\nredundant operations on the runtime's part. However, this gives us a\r\nbetter coverage and fixes the issue as a short-term solution.\r\n\r\nIdeally, instead of recording \"pid gets\" we should cache array/domain\r\nrecords accessed in the loop body in a set. And then generate \"pid gets\"\r\nin the preamble using that set.\r\n\r\n[Merging to fix a nightly failure, pending post-merge review]\r\n\r\nTest:\r\n- [x] nvidia\r\n- [x] nvidia `--gpu-specialization` in `gpu/native/distArray`","shortMessageHtmlLink":"GPU Privatization: handle a case where array record is LICM'd (chapel…"}},{"before":"18378330c074ebbf0be27d867a53183697f20532","after":"9cd20851fd1248691beceb679e1b56a592e4f5c4","ref":"refs/heads/main","pushedAt":"2024-02-07T19:26:22.000Z","pushType":"push","commitsCount":147,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"Update Arkouda release testing to 1.33 (#24355)\n\n[ trivial, not reviewed ]","shortMessageHtmlLink":"Update Arkouda release testing to 1.33 (chapel-lang#24355)"}},{"before":"d4d76e316f110f0079f7d81c351042185a2a26d4","after":"18378330c074ebbf0be27d867a53183697f20532","ref":"refs/heads/main","pushedAt":"2024-02-02T15:10:35.000Z","pushType":"push","commitsCount":105,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"'domainDistIsLayout' --> 'chpl_domainDistIsLayout' (#24317)\n\nThis PR adds the `chpl_` prefix to `domainDistIsLayout()`\r\nin order to remove it from the public API.\r\n\r\ndomainDistIsLayout() was made public in #18829 so it can be used\r\nacross the internal modules. We consider such addition a bug\r\nbecause this function has not received a design review\r\nand was/is not intended to be user-facing. This PR fixes that bug.\r\n\r\nThis also removes the intended-for-user comment added to\r\ndomainDistIsLayout in #23664, as it is no longer needed.\r\n\r\nr: @dlongnecke-cray","shortMessageHtmlLink":"'domainDistIsLayout' --> 'chpl_domainDistIsLayout' (chapel-lang#24317)"}},{"before":"366887b505d41a137af18e3a3945ff17f71d88aa","after":"d4d76e316f110f0079f7d81c351042185a2a26d4","ref":"refs/heads/main","pushedAt":"2024-01-30T17:09:10.000Z","pushType":"push","commitsCount":92,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"Fix test prediff to use more portable sed (#24298)\n\nFixes test using sed which failed in Mac nightly testing since the `sed`\r\ncommand did not match.\r\n\r\nTested on Mac and Linux\r\n\r\n[Not reviewed - trivial]","shortMessageHtmlLink":"Fix test prediff to use more portable sed (chapel-lang#24298)"}},{"before":"84232c69a1745f48816560062334373ef1b1a52f","after":"366887b505d41a137af18e3a3945ff17f71d88aa","ref":"refs/heads/main","pushedAt":"2024-01-27T20:06:08.000Z","pushType":"push","commitsCount":56,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"LSP: Shim Chapel to generate compiler commands; share Contexts between related files. (#24254)\n\nThis PR adds a new script called 'shim-chpl', which makes it possible to\r\nfigure out the configuration in which Chapel was invoked. This way, we\r\ncan tell the language server which additional flags to provide to Dyno,\r\nand thus to resolve projects such as Arkouda.\r\n\r\nFor instance, instead of just compiling Arkouda:\r\n\r\n```Bash\r\nmake server\r\n```\r\n\r\nWe can preface the command with `shim-chpl`:\r\n\r\n```Bash\r\nshim-chpl make server\r\n```\r\n\r\nThis runs the exact same commands as `make server`, but additionally\r\nnotes the compilation configuration in a `.cls-info.json` file.\r\n\r\n```JSON\r\n{\r\n \"B.chpl\": [{\"module_dirs\": [\"added_modules\"], \"files\": [\"B.chpl\", \"A.chpl\"]}],\r\n \"A.chpl\": [{\"module_dirs\": [\"added_modules\"], \"files\": [\"B.chpl\", \"A.chpl\"]}]\r\n}\r\n```\r\n\r\nThe language server is then made to load and process the\r\n`.cls-info.json` files. From these files, we are able to learn enough\r\ninformation to resolve projects with a non-flat module directory, like\r\nArkouda.\r\n\r\n![\"Screen]()
\r\n\r\nIn the process of making this work, I've also applied a major\r\narchitectural change: moving all global state into a\r\n`ChapelLanguageServer` class that extends from `pygls`'s\r\n`LanguageServer`. This made it possible to clean up some code, as well\r\nas to avoid situations in which somehow two language servers could be\r\ntouching the same state.\r\n\r\nReviewed by @jabraham17 -- thanks!","shortMessageHtmlLink":"LSP: Shim Chapel to generate compiler commands; share Contexts betwee…"}},{"before":"bd916b21688a258a0013f9c5665bcd6eeb5d2a19","after":"84232c69a1745f48816560062334373ef1b1a52f","ref":"refs/heads/main","pushedAt":"2024-01-25T15:12:18.000Z","pushType":"push","commitsCount":414,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"Load deps from Spack install on Osprey (#24271)\n\nSwitch to loading Chapel nightly test dependencies via new Spack install\r\nfor Osprey.\r\n\r\n[reviewed by @arezaii , thanks!]\r\n\r\nTesting:\r\n- [x] a few manual test runs on Osprey with the change","shortMessageHtmlLink":"Load deps from Spack install on Osprey (chapel-lang#24271)"}},{"before":"8ede09f2eb49950c0f8f39fb9cffb7d01be71a05","after":"bd916b21688a258a0013f9c5665bcd6eeb5d2a19","ref":"refs/heads/main","pushedAt":"2024-01-12T05:46:58.000Z","pushType":"push","commitsCount":166,"pusher":{"login":"amritpan","name":"Amritpan Kaur","path":"/amritpan","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/32915938?s=80&v=4"},"commit":{"message":"--print-additional-errors : another use (#24163)\n\nThis PR enables the existing compiler option `--print-additional-errors`\r\nto print errors that are being postponed, ex., when resolving copy\r\ninitialization, such as chpl__coerceCopy and chpl__coerceMove,\r\nas in `var B = (an array slice);`.\r\n\r\nThe benefit is that this enables printing the call stack leading to such\r\nan error. When a postponed error is finally printed, if at all,\r\nthe call stack is no longer available. Without a call stack,\r\nit is often impossible to detect the cause of the error\r\nwithout running the compiler under the debugger.\r\n\r\nNext steps: have the compiler recommend compiling with\r\n`--print-additional-errors` when this can explain the final compilation error.\r\n\r\nr: @jabraham17","shortMessageHtmlLink":"--print-additional-errors : another use (chapel-lang#24163)"}}],"hasNextPage":false,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"Y3Vyc29yOnYyOpK7MjAyNC0wMy0xOFQyMToyMjo1OS4wMDAwMDBazwAAAAQY_UCN","startCursor":"Y3Vyc29yOnYyOpK7MjAyNC0wMy0xOFQyMToyMjo1OS4wMDAwMDBazwAAAAQY_UCN","endCursor":"Y3Vyc29yOnYyOpK7MjAyNC0wMS0xMlQwNTo0Njo1OC4wMDAwMDBazwAAAAPd545n"}},"title":"Activity · amritpan/chapel"}