Feed Advanced Search

Sun, Feb 4

ezyang created D4376: Restore 'It is a member of hidden package' message..
Sun, Feb 4, 5:45 PM

Dec 1 2017

ezyang committed rGHC6998772043a7: Make use of boot TyThings during typechecking. (authored by ezyang).
Make use of boot TyThings during typechecking.
Dec 1 2017, 7:38 AM
ezyang committed rGHCe1fb28384c44: Handle CPP properly in Backpack (authored by ezyang).
Handle CPP properly in Backpack
Dec 1 2017, 7:38 AM
ezyang closed D4234: Handle CPP properly in Backpack.
Dec 1 2017, 7:38 AM
ezyang closed D4154: Make use of boot TyThings during typechecking..
Dec 1 2017, 7:38 AM

Nov 24 2017

ezyang updated the diff for D4155: Don't complain about UNPACK in -fno-code..

make this actually work

Nov 24 2017, 11:20 PM
ezyang added a comment to D4154: Make use of boot TyThings during typechecking..

There are a lot of unknowns with this patch (e.g., whether or not it will cause some subtle bugs because a TyThing is not wired up) but Simon and I are in agreement that it's worth a try. Unfortunately harbormaster is segfaulting right now but I know this patch validate(d) on OS X without any trouble.

Nov 24 2017, 11:15 PM
ezyang created D4234: Handle CPP properly in Backpack.
Nov 24 2017, 11:03 PM

Nov 23 2017

ezyang added a comment to D4155: Don't complain about UNPACK in -fno-code..

This seems great to me, but then I don't understand why -O0 suppressed this originally.

Nov 23 2017, 10:02 AM
ezyang requested changes to D4208: Ensure that -M mode accounts for boot file dependencies.

Still waiting for the test :) The patch looks OK, though.

Nov 23 2017, 9:58 AM

Nov 17 2017

ezyang updated the diff for D4154: Make use of boot TyThings during typechecking..

one more validate fix

Nov 17 2017, 12:34 AM

Nov 16 2017

ezyang updated the diff for D4155: Don't complain about UNPACK in -fno-code..

Tweak logic

Nov 16 2017, 9:26 PM
ezyang updated the diff for D4154: Make use of boot TyThings during typechecking..

validate clean

Nov 16 2017, 9:02 PM
ezyang requested changes to D4159: ghc-pkg: recompute `abi-depends` for updated packages.

I think this patch is a strict improvement over the current situation, but I don't think it is correct: see my comment about package database stacks.

Nov 16 2017, 5:56 PM
ezyang abandoned D4138: New strategy for deferring type family consistency checks..

We're going to go the route of the other revision.

Nov 16 2017, 5:40 PM
ezyang added a comment to D4155: Don't complain about UNPACK in -fno-code..

OK, in this case we need to flip this patch on its head. Today, if you compile with -O0, GHC will not warn about ill-placed UNPACK pragmas. We should make this be the case. And then I need to add some more logic to have this warning be suppressed when Backpack is being used,.

Nov 16 2017, 5:40 PM

Nov 5 2017

ezyang created D4155: Don't complain about UNPACK in -fno-code..
Nov 5 2017, 2:18 PM
ezyang created D4154: Make use of boot TyThings during typechecking..
Nov 5 2017, 2:06 PM

Oct 30 2017

ezyang updated the diff for D4138: New strategy for deferring type family consistency checks..

fix validate failures

Oct 30 2017, 9:47 AM

Oct 29 2017

ezyang created D4138: New strategy for deferring type family consistency checks..
Oct 29 2017, 8:40 PM

Oct 17 2017

ezyang added a comment to D4102: Error when deriving instances in hs-boot files.

Don't forget to check in the expected stderr.

Oct 17 2017, 9:23 PM

Oct 12 2017

ezyang added a comment to D3836: Add module ParUpsweep, and leverage it in GhcMake.parUpsweep to enable better parallelism.

I mentioned this changeset to simonpj, and his thought was 12% perf improvement was not all that much for a decent increase in code complexity (2x speed up would be much better). Mentioning it on this ticket so I don't forget.

Oct 12 2017, 2:04 PM

Oct 8 2017

ezyang created D4081: Give a reference to Foreign.Concurrent..
Oct 8 2017, 7:19 PM

Oct 7 2017

ezyang updated the diff for D3825: Levity polymorphic Backpack..

don't print kinds when they are boring

Oct 7 2017, 12:12 PM

Oct 4 2017

ezyang updated the diff for D3825: Levity polymorphic Backpack..

rebase

Oct 4 2017, 1:56 PM

Oct 3 2017

ezyang updated the diff for D3825: Levity polymorphic Backpack..

rebase for ci

Oct 3 2017, 4:11 PM

Oct 1 2017

ezyang updated the diff for D3825: Levity polymorphic Backpack..

try again ci

Oct 1 2017, 11:27 PM
ezyang created D4057: Include libraries which fill holes as deps when linking..
Oct 1 2017, 10:07 PM

Sep 30 2017

ezyang requested changes to D3836: Add module ParUpsweep, and leverage it in GhcMake.parUpsweep to enable better parallelism.

Had an offline conversation with duog about this patch. There is some delicate coroutining going on between the main scheduler thread of the new parallel compilation phase, and the "yield" mechanism inside HscMain. There should be a note on this.

Sep 30 2017, 4:54 PM
ezyang requested changes to D3832: Change HscEnv.hsc_HPT from HomePackageTable to IORef HomePackageTable.

duog and I chatted about this offline, and we came up with a scenario where this change is unsound.

Sep 30 2017, 4:49 PM

Sep 26 2017

ezyang added a comment to D3836: Add module ParUpsweep, and leverage it in GhcMake.parUpsweep to enable better parallelism.

I'm not done reviewing, but let me add some initial comments.

Sep 26 2017, 1:53 PM

Sep 20 2017

ezyang added a comment to D3898: [RFC] Introduce -ddump-mod-graph.

I like to cite the README here: https://github.com/ndmitchell/ghc-make for demonstrating why *anything* is better than what we have in ghc --make today. It's pretty simple, really: ghc --make *always* parses the module headers of every module in your package, no ifs ands or buts. A shake based approach can do a single database load, and a quick probe of the timestamps of all files, before deciding no recompilation must be done.

Sep 20 2017, 7:47 PM

Aug 29 2017

ezyang added a comment to D3892: desugar: Ensure that a module's dep_orphs doesn't contain itself.

Well, this clearly fixes a bug, so I don't mind taking it. But I feel like we have similar logic for this (filter out self import) elsewhere in GHC, so I wonder if we've hit the root cause.

Aug 29 2017, 7:55 PM

Aug 24 2017

ezyang added a comment to D3825: Levity polymorphic Backpack..

It... seems like I have managed to convince Richard that this patch won't blow things up... so I guess we can merge it :)

Aug 24 2017, 9:19 PM
ezyang accepted D3864: Remember the AvailInfo for each IE.

I like the new approach!

Aug 24 2017, 9:19 PM
ezyang added a comment to D3832: Change HscEnv.hsc_HPT from HomePackageTable to IORef HomePackageTable.

It would be easier to give a judgment on HPT in an IORef if I knew what design forces conspired to put you into this regime. This particular change comes with the risk that some callers are hanging on to HPT because they believe it won't update, when it actually does.

Aug 24 2017, 8:45 PM

Aug 13 2017

ezyang added a comment to D3836: Add module ParUpsweep, and leverage it in GhcMake.parUpsweep to enable better parallelism.

It will take me some time to review this patch, but there are two things I would think about. First, consider using threadscope or a similar tool to get traces showing how long parsing/typechecking/codegen take as "bars" (think similar to nvrtc; Google it) which will help you get a qualitative perspective on how much your patch is helping. If you need a machine with more cores to test on I can volunteer one. Second, take care on whether or not all the thunks are being forced between stages, or if you're actually foisting work into a later phase (which can defeat parallelization.)

Aug 13 2017, 12:00 PM
ezyang added a comment to D3844: RFC: TcRnExports remembers `AvailInfo` for each IE with -haddock.

I feel like this commit could be made more type-safe, since there seems to be some implicit invariant on how IEAvails is supposed to be inserted in IEs. I'm sympathetic to the desire to avoid API breaking changes but I think doing it properly is it worth it here.

Aug 13 2017, 11:15 AM

Aug 7 2017

ezyang added a comment to D3825: Levity polymorphic Backpack..

Oh! Rep is going to be a AlgTyCon, with an AbstractTyCon RHS.

Aug 7 2017, 8:58 AM

Aug 6 2017

ezyang added a comment to D3825: Levity polymorphic Backpack..

Can such a module be compiled?

Aug 6 2017, 8:52 PM

Aug 5 2017

ezyang created D3825: Levity polymorphic Backpack..
Aug 5 2017, 11:38 PM
ezyang added a comment to D3815: Fix incorrect retypecheck loop in -j (#14075).

@duog, you are right. I have written a ticket about the issue:
https://ghc.haskell.org/trac/ghc/ticket/14092#ticket

Aug 5 2017, 10:19 PM

Aug 1 2017

ezyang created D3815: Fix incorrect retypecheck loop in -j (#14075).
Aug 1 2017, 11:51 PM

Jul 18 2017

ezyang updated the diff for D3742: Make IfaceAxiom typechecking lazier..

more comments

Jul 18 2017, 7:41 PM

Jul 15 2017

ezyang created D3743: Fix hs-boot knot-tying with record wild cards..
Jul 15 2017, 7:00 PM
ezyang created D3742: Make IfaceAxiom typechecking lazier..
Jul 15 2017, 6:23 PM

Jun 14 2017

ezyang added a comment to D3647: Introduce module hierarchy.

I'm a grumpy curmudgeon and don't want to have to learn new locations for modules, but don't let me stop you :) One thing that would be kinder to all GHC API users (who are now going to have to rewrite all of their imports) is to create an old-ghc package which reexports all of the modules under their old names. Might be helpful.

Jun 14 2017, 9:51 PM

May 16 2017

ezyang committed rGHCd9e9a9b3016a: Fix #13703 by correctly using munged names in ghc-pkg. (authored by ezyang).
Fix #13703 by correctly using munged names in ghc-pkg.
May 16 2017, 9:00 PM
ezyang closed D3590: Fix #13703 by correctly using munged names in ghc-pkg. by committing rGHCd9e9a9b3016a: Fix #13703 by correctly using munged names in ghc-pkg..
May 16 2017, 9:00 PM

May 15 2017

ezyang created D3590: Fix #13703 by correctly using munged names in ghc-pkg..
May 15 2017, 11:22 PM
ezyang added inline comments to D3588: Cache conversion to a map for dep_mods.
May 15 2017, 7:46 PM

May 13 2017

ezyang abandoned D3514: Revert "Fix #10923 by fingerprinting optimization level.".

OK! Onwards to a brave new future :)

May 13 2017, 8:53 AM

Apr 30 2017

ezyang added a comment to D3514: Revert "Fix #10923 by fingerprinting optimization level.".

I can't commit to figuring out how to deal with the regression (Trac #13604) in time for 8.2, so better to leave things the way they were previously :/

Apr 30 2017, 3:11 PM
ezyang created D3514: Revert "Fix #10923 by fingerprinting optimization level.".
Apr 30 2017, 1:43 PM

Apr 25 2017

ezyang created D3501: Update Cabal submodule, with necessary wibbles..
Apr 25 2017, 8:42 PM

Apr 22 2017

ezyang committed rGHC6cffee6a567a: Haddock submodule update. (authored by ezyang).
Haddock submodule update.
Apr 22 2017, 10:42 PM

Apr 11 2017

ezyang added a comment to D3441: Compile modules that are needed by template haskell, even with -fno-code..

So, to explain the history behind -fwrite-interface, it was added (ab105f83dcd5f9094a9edb0f0c8266fba6f3c808) because I noticed that -fno-code wasn't writing out interface files, and this meant that recompilation didn't work in "flycheck" mode because you couldn't avoid retypechecking a file that hadn't changed. I added a new flag rather than change the behavior of -fno-code because it was (a) more convenient from an implementation perspective, and (b) preserved BC (for arguably bad behavior.)

Apr 11 2017, 6:30 PM
ezyang added a comment to D3441: Compile modules that are needed by template haskell, even with -fno-code..

I lean towards interpreter (because I want -fno-code to run as quickly as possible), but what seems important to me is the ability to toggle between the two, similarly to how -fobject-code can be used to force GHCi to asm rather than interpret.

Apr 11 2017, 2:47 PM

Apr 6 2017

ezyang added a comment to D1091: Handle annotations in .hsig files (fixes #10621).

It basically needs to be rewritten for the new Backpack stuff.

Apr 6 2017, 8:43 PM
ezyang added a comment to D3318: GHC_STAGE1 isn't defined, use other form..

No, but I landed the fix that worked for me, see 5db415580e0738f934e35b7012fe35a79b7e97c7

Apr 6 2017, 6:13 PM

Apr 3 2017

ezyang committed rGHCf541fc63060a: Fix validate. (authored by ezyang).
Fix validate.
Apr 3 2017, 4:24 AM

Apr 2 2017

ezyang committed rGHC852a43f360af: Correctly handle wired in unit IDs in -instantiated-with (authored by ezyang).
Correctly handle wired in unit IDs in -instantiated-with
Apr 2 2017, 6:50 PM
ezyang closed D3385: Correctly handle wired in unit IDs in -instantiated-with by committing rGHC852a43f360af: Correctly handle wired in unit IDs in -instantiated-with.
Apr 2 2017, 6:50 PM
ezyang committed rGHC45d33f35f689: Better test coverage for module reexports in signatures. (authored by ezyang).
Better test coverage for module reexports in signatures.
Apr 2 2017, 6:50 PM
ezyang committed rGHC0c333c89ffce: Extra docs on tcg_imports. (authored by ezyang).
Extra docs on tcg_imports.
Apr 2 2017, 6:49 PM
ezyang committed rGHC71dadd71e7bb: Extra docs on exports_from_avail. (authored by ezyang).
Extra docs on exports_from_avail.
Apr 2 2017, 6:49 PM
ezyang committed rGHCd2df718cfb9d: Add more documentation on mergeSignatures. (authored by ezyang).
Add more documentation on mergeSignatures.
Apr 2 2017, 6:49 PM
ezyang committed rGHC5db415580e07: mk/boilerplate.mk defines STAGE1_GHC, not GHC_STAGE1. (authored by ezyang).
mk/boilerplate.mk defines STAGE1_GHC, not GHC_STAGE1.
Apr 2 2017, 6:49 PM
ezyang committed rGHC5fb485a3e1e8: Fix recompilation avoidance bug for implementor of hsig. (authored by ezyang).
Fix recompilation avoidance bug for implementor of hsig.
Apr 2 2017, 6:49 PM
ezyang closed D3381: Fix recompilation avoidance bug for implementor of hsig. by committing rGHC5fb485a3e1e8: Fix recompilation avoidance bug for implementor of hsig..
Apr 2 2017, 6:49 PM
ezyang committed rGHCd4e8ebcd04cc: Minor comment updates on CSE. (authored by ezyang).
Minor comment updates on CSE.
Apr 2 2017, 6:49 PM

Mar 31 2017

ezyang added a comment to D3392: Optimise common cases of GHC.setProgramDynFlags.

Well, that's the problem, I don't really know how to do that; all I know is that I change my environment file and ask GHC to "reload its configuration", it would be surprising if it didn't get picked up...

Mar 31 2017, 9:52 AM

Mar 30 2017

ezyang added a comment to D3392: Optimise common cases of GHC.setProgramDynFlags.

I don't think the extraPkgConfs test is enough. If someone applies removeGlobalPkgConf we won't notice it if we run the function on the empty list. I guess the solution here is to get an actual data type (I never liked the function anyway.)

Mar 30 2017, 2:43 PM

Mar 29 2017

ezyang updated the diff for D3381: Fix recompilation avoidance bug for implementor of hsig..

Better cleaner implementation

Mar 29 2017, 1:05 PM
ezyang requested changes to D3392: Optimise common cases of GHC.setProgramDynFlags.
Mar 29 2017, 11:51 AM

Mar 27 2017

ezyang abandoned D3253: Fix roles merging to apply only to non-rep-injective types..

This got merged (but the commit message was truncated so Phabricator didn't see it.)

Mar 27 2017, 1:57 PM

Mar 26 2017

ezyang added inline comments to D3318: GHC_STAGE1 isn't defined, use other form..
Mar 26 2017, 4:39 PM
ezyang updated the diff for D3381: Fix recompilation avoidance bug for implementor of hsig..

fix validate error

Mar 26 2017, 4:30 PM
ezyang created D3385: Correctly handle wired in unit IDs in -instantiated-with.
Mar 26 2017, 4:08 PM
ezyang created D3381: Fix recompilation avoidance bug for implementor of hsig..
Mar 26 2017, 12:54 AM

Mar 22 2017

ezyang committed rGHCacd85ce97acc: Haddock submodule update. (authored by ezyang).
Haddock submodule update.
Mar 22 2017, 4:15 PM

Mar 20 2017

ezyang committed rGHCe0eaea918c32: Correctly account for -package-db ordering when picking packages. (authored by ezyang).
Correctly account for -package-db ordering when picking packages.
Mar 20 2017, 8:18 PM
ezyang closed D3369: Correctly account for -package-db ordering when picking packages. by committing rGHCe0eaea918c32: Correctly account for -package-db ordering when picking packages..
Mar 20 2017, 8:18 PM
ezyang committed rGHC40b65db4cd34: Document the perplexing reversed nature of extraPkgConfs and friends. (authored by ezyang).
Document the perplexing reversed nature of extraPkgConfs and friends.
Mar 20 2017, 8:18 PM
ezyang committed rGHC09485bba2b9a: UniqMap implementation. (authored by ezyang).
UniqMap implementation.
Mar 20 2017, 8:18 PM

Mar 19 2017

ezyang created D3369: Correctly account for -package-db ordering when picking packages..
Mar 19 2017, 6:15 PM

Mar 17 2017

ezyang added a comment to D3318: GHC_STAGE1 isn't defined, use other form..

I reverted it. I'll look more carefully why the boilerplate isn't being applied.

Mar 17 2017, 9:38 PM
ezyang committed rGHCad191042896c: Revert "GHC_STAGE1 isn't defined, use other form." (authored by ezyang).
Revert "GHC_STAGE1 isn't defined, use other form."
Mar 17 2017, 9:36 PM
ezyang added a reverting commit for rGHC138434fbef32: GHC_STAGE1 isn't defined, use other form.: rGHCad191042896c: Revert "GHC_STAGE1 isn't defined, use other form.".
Mar 17 2017, 9:36 PM
ezyang committed rGHC138434fbef32: GHC_STAGE1 isn't defined, use other form. (authored by ezyang).
GHC_STAGE1 isn't defined, use other form.
Mar 17 2017, 3:48 AM
ezyang closed D3318: GHC_STAGE1 isn't defined, use other form. by committing rGHC138434fbef32: GHC_STAGE1 isn't defined, use other form..
Mar 17 2017, 3:48 AM
ezyang committed rGHCb301f78eee89: Save renamed syntax when signature merging. (authored by ezyang).
Save renamed syntax when signature merging.
Mar 17 2017, 3:48 AM
ezyang closed D3335: Save renamed syntax when signature merging. by committing rGHCb301f78eee89: Save renamed syntax when signature merging..
Mar 17 2017, 3:48 AM
ezyang created D3357: Document hithertoo undocumented HPCTIXFILE option..
Mar 17 2017, 1:24 AM

Mar 13 2017

ezyang added inline comments to D3335: Save renamed syntax when signature merging..
Mar 13 2017, 4:16 PM
ezyang updated the diff for D3335: Save renamed syntax when signature merging..

add more comments

Mar 13 2017, 4:16 PM
ezyang created D3335: Save renamed syntax when signature merging..
Mar 13 2017, 4:22 AM

Mar 10 2017

ezyang added inline comments to D3318: GHC_STAGE1 isn't defined, use other form..
Mar 10 2017, 6:01 PM
ezyang added inline comments to D3318: GHC_STAGE1 isn't defined, use other form..
Mar 10 2017, 5:38 PM
ezyang created D3318: GHC_STAGE1 isn't defined, use other form..
Mar 10 2017, 4:34 PM