- Oct 20, 2020
-
-
Sylvain Henry authored
-
- Oct 09, 2020
-
-
Sylvain Henry authored
-
Sylvain Henry authored
-
Sylvain Henry authored
Every Core/Stg case-expression now carries a set of variable names to keep alive during the evaluation of the scrutinee. During StgToCmm for case-expression, we now generate an additional `touch#` pseudo-instruction call for every variable in the keep-alive set after the evaluation of the scrutinee. This approach should allow us to fix issues with the current `touch#` primop (exposed to user Haskell code) which is not correctly taken into account by the simplifier leading to segfaults (#17746, #18040, #18061).
-
- Oct 07, 2020
-
-
* Include funTyCon in exposedPrimTyCons. Every single place using exposedPrimTyCons was adding funTyCon manually. * Remove unused synTyConResKind and ieLWrappedName * Add recordSelectorTyCon_maybe * In exprType, panic instead of giving a trace message and dummy output. This prevents #18767 reoccurring. * Fix compilation error in fragile concprog001 test (part of #18732)
-
Several people have struggled with metric change annotations in their commit messages not being recognized due to the fact that GitLab's job log inserts a space at the beginning of each line. Teach the regular expression to accept this whitespace.
-
-
When linear types are disabled, HsUnrestrictedArrow is treated as HslinearArrow. Move this adjustment into the type checking phase, so that the parsed source accurately represents the source as parsed. Closes #18791
-
-
Before this patch -flink-rts could link with GHC's rts instead of the selected one.
-
-
Move the atomix exchange over the Ptr type to an internal module. Fix a bug caused by us passing ptr-to-ptr instead of ptr to atomic exchange. Renamed interlockedExchange to exchangePtr. I've also added an cas primitive. It turned out we don't need it for WinIO but I'm leaving it in as it's useful for other things.
-
-
- Oct 05, 2020
-
-
Previously `TICK_BUMP_BY` was defined as ```c #define TICK_BUMP_BY(ctr,n) CLong[ctr] = CLong[ctr] + n ``` Yet the tickers themselves were defined as `StgInt`s. This happened to work out correctly on Linux, where `CLong` is 64-bits. However, it failed on Windows, where `CLong` is 32-bits, resulting in #18782. Fixes #18783.
-
-
Currently, `integerDecodeDouble#` is known-key so that it can be recognised in constant folding. But that is very brittle and doesn't survive worker/wrapper, which we even do for `NOINLINE` things since #13143. Also it is a trade-off: The implementation of `integerDecodeDouble#` allocates an `Integer` box that never cancels aways if we don't inline it. Hence we recognise the `decodeDouble_Int64#` primop instead in constant folding, so that we can inline `integerDecodeDouble#`. As a result, `integerDecodeDouble#` no longer needs to be known-key. While doing so, I realised that we don't constant-fold `decodeFloat_Int#` either, so I also added a RULE for it. `integerDecodeDouble` is dead, so I deleted it. Part of #18092. This improves the 32-bit `realToFrac`/`toRational`: Metric Decrease: T10359
-
- Oct 03, 2020
-
-
-
- Fix formatting of code blocks and a few sphinx warnings - Move the Void# change to 9.2, it was done right after the branch was cut - Fix typo in linear types documentation - Note that -Wincomplete-uni-patterns affects lazy patterns [skip ci]
-
- Oct 02, 2020
-
-
Patch taken from #18624 (comment 300673)
-
Incidentally fix powModInteger which was crashing in integer-gmp for negative exponents when the modular multiplicative inverse for the base didn't exist. Now we compute it explicitly with integerRecipMod so that every backend returns the same result without crashing.
-
-
1. Fix and update section headers in GHC/Hs/Extension.hs 2. Delete the unused 'XCoreAnn' and 'XTickPragma' families 3. Avoid calls to 'panic' in 'pprStmt'
-
-
[skip ci]
-
- Oct 01, 2020
-
-
It did not do any useful work.
-
-
A follow-up to !4020 (5830a12c)
-
Close #18755
-
Because the generated `KindRep`s don't have an unfolding, !3230 did not actually stop to compute, attach and serialise unnecessary CPR signatures for them. As already said in `Note [CPR for data structures]`, that leads to bloated interface files which is ultimately quadratic for Nested CPR. So we don't attach any CPR signature to bindings that * Are not thunks (because thunks are not in WHNF) * Have arity 0 (which means the top-level constructor is not a lambda) If the data structure has an unfolding, we continue to look through it. If not (as is the case for `KindRep`s), we look at the unchanged CPR signature and see `topCprType`, as expected.
-
-
These are not reported by GHC because Happy adds {-# OPTIONS_GHC -w #-} -
-
Haskell and Cmm parsers/lexers now report errors and warnings using ADTs defined in GHC.Parser.Errors. They can be printed using functions in GHC.Parser.Errors.Ppr. Some of the errors provide hints with a separate ADT (e.g. to suggest to turn on some extension). For now, however, hints are not consistent across all messages. For example some errors contain the hints in the main message. I didn't want to change any message with this patch. I expect these changes to be discussed and implemented later. Surprisingly, this patch enhances performance. On CI (x86_64/deb9/hadrian, ghc/alloc): parsing001 -11.5% T13719 -2.7% MultiLayerModules -3.5% Naperian -3.1% Bump haddock submodule Metric Decrease: MultiLayerModules Naperian T13719 parsing001 -
-
-
Add `mainModuleNameIs` to DynFlags and demote `mainModIs` to function which uses the homeUnit from DynFlags it is created from.
-
Metric Decrease: T18574
-
-
Ben Gamari authored
-
- Sep 30, 2020
-
-
Close #10709
-
