- May 05, 2022
-
-
-
-
-
-
IntMap.size is O(n). The new code should be slightly more efficient. The transformation of GHC.CmmToAsm.CFG.calcFreqs.nodeCount can be described formally as the transformation: (\sum_{0}^{n-1} \sum_{0}^{k-1} i_nk) + n ==> (\sum_{0}^{n-1} 1 + \sum_{0}^{k-1} i_nk) -
-
-
- May 04, 2022
-
-
Windows binary distributions built with Hadrian have a target platform suffix in the name of their root directory. Teach `ci.sh` about this fact. (cherry picked from commit df5752f3)
-
(cherry picked from commit 16d6a8ff)
-
It appears that the version of sphinx shipped on CentOS 7 reports a version string of `Sphinx v1...`. Accept the `v`. (cherry picked from commit a9197a29)
-
(cherry picked from commit fd08b0c9)
-
Due to #16177. Also cleanup some code style issues. (cherry picked from commit cc1c3861)
-
Darwin requires the `-r` flag to be compatible with GNU sed. (cherry picked from commit 512338c8)
-
Pull forward minimum version to match 9.2. (cherry picked from commit c26faa54)
-
-
Our error reporting in generated code (via desugaring before typechecking) only worked when the generated code was just a simple call. This commit makes it work in nested cases.
-
Previously the make build system unconditionally included StgCRunAsm.S in the link, meaning that the RTS would require an execstack unnecessarily. Fixes #21478.
-
The LaTeX documentation generator does not seem to have been used for quite some time, so the LaTeX-to-Haddock preprocessing step has become a pointless complication that makes documenting the contents of GHC.Prim needlessly difficult. This commit replaces the LaTeX syntax with the Haddock it would have been converted into, anyway, though with an additional distinction: it uses single quotes in places to instruct Haddock to generate hyperlinks to bindings. This improves the quality of the generated output.
-
-
For expressions like `(scc<cc_name> primOp#) arg1` we should also look at arg1 to determine if we call primOp# at a fixed runtime rep. This is what corePrep already does but CoreLint didn't yet. This patch will bring them in sync in this regard. It also uses tickishFloatable in CorePrep instead of CorePrep having it's own slightly differing definition of when a tick is floatable.
-
- May 03, 2022
-
-
Sebastian Graf authoredSee the new `Note [SubDemand denotes at least one evaluation]`. A demand `n :* sd` on a let binder `x=e` now means > "`x` was evaluated `n` times and in any program trace it is evaluated, `e` is > evaluated deeply in sub-demand `sd`." The "any time it is evaluated" premise is what this patch adds. As a result, we get better nested strictness. For example (T21081) ```hs f :: (Bool, Bool) -> (Bool, Bool) f pr = (case pr of (a,b) -> a /= b, True) -- before: <MP(L,L)> -- after: <MP(SL,SL)> g :: Int -> (Bool, Bool) g x = let y = let z = odd x in (z,z) in f y ``` The change in demand signature "before" to "after" allows us to case-bind `z` here. Similarly good things happen for the `sd` in call sub-demands `Cn(sd)`, which allows for more eta-reduction (which is only sound with `-fno-pedantic-bottoms`, albeit). We also fix #21085, a surprising inconsistency with `Poly` to `Call` sub-demand expansion. In an attempt to fix a regression caused by less inlining due to eta-reduction in T15426, I eta-expanded the definition of `elemIndex` and `elemIndices`, thus fixing #21345 on the go. The main point of this patch is that it fixes #21081 and #21133. Annoyingly, I discovered that more precise demand signatures for join points can transform a program into a lazier program if that join point gets floated to the top-level, see #21392. There is no simple fix at the moment, but !5349 might. Thus, we accept a ~5% regression in `MultiLayerModulesTH_OneShot`, where #21392 bites us in `addListToUniqDSet`. T21392 reliably reproduces the issue. Surprisingly, ghc/alloc perf on Windows improves much more than on other jobs, by 0.4% in the geometric mean and by 2% in T16875. Metric Increase: MultiLayerModulesTH_OneShot Metric Decrease: T16875
-
- May 02, 2022
-
-
-
Note [Nullary unboxed tuple] was removed in e9e61f18. This codepath is tested by T15696_3.
-
Fixes ghc/ghc#21420
-
-
We used to check the divergence and that the number of arguments > arity. But arity zero represents unknown arity so this was subtly broken for a long time! We would check if the saturated function diverges, and if we applied >=arity arguments. But for unknown arity functions any number of arguments is >=idArity. This fixes #21440.
-
close #21417
-
We use compatibleRep to compare reps, and avoid checking functions with levity polymorphic types because of #21399.
-
- Apr 30, 2022
-
-
-
This fixes an assertion failure in the m32 allocator due to the imprecisely specified preconditions of `m32_allocator_push_filled_list`. Specifically, the caller must ensure that the page type is set to filled prior to calling `m32_allocator_push_filled_list`. While this issue did result in an assertion failure in the debug RTS, the issue is in fact benign.
-
There have been a number of tickets about non-tested flavours not passing the testsuite.. this is expected and now noted in the documentation. You use other flavours to run the testsuite at your own risk. Fixes #21418
-
-
This reverts commit ef013593. See ticket #21229 ------------------------- Metric Decrease: T15164 Metric Increase: T13056 -------------------------
-
-
-
-
Change the rewrite rule examples to include a space between the composition of `f` and `g` in the map rewrite rule examples. Without this change, if the user has locally enabled the extension OverloadedRecordDot the copied example will result in a compile time error that `g` is not a field of `f`. ``` • Could not deduce (GHC.Records.HasField "g" (a -> b) (a1 -> b)) arising from selecting the field ‘g’ ``` -
Replaces uses of `TcRnUnknownMessage` with proper diagnostics constructors.
-
Previously, whenever `mkFunCo` would produce reflexive coercions, it would use `mkVisFunTy` to produce the kind of the coercion. However, `mkFunCo` is also used to produce coercions between types of the form `ty1 => ty2` in certain places. This has the unfortunate side effect of causing the type of the coercion to appear as `ty1 -> ty2` in certain error messages, as spotted in #21328. This patch address this by changing replacing the use of `mkVisFunTy` with `mkFunctionType` in `mkFunCo`. `mkFunctionType` checks the kind of `ty1` and makes the function arrow `=>` instead of `->` if `ty1` has kind `Constraint`, so this should always produce the correct `AnonArgFlag`. As a result, this patch fixes part (2) of #21328. This is not the only possible way to fix #21328, as the discussion on that issue lists some possible alternatives. Ultimately, it was concluded that the alternatives would be difficult to maintain, and since we already use `mkFunctionType` in `coercionLKind` and `coercionRKind`, using `mkFunctionType` in `mkFunCo` is consistent with this choice. Moreover, using `mkFunctionType` does not regress the performance of any test case we have in GHC's test suite.
-
In principle, the *visible* instances are * all instances defined in a prior top-level declaration group (see docs on `newDeclarationGroup`), or * all instances defined in any module transitively imported by the module being compiled However, actually searching all modules transitively below the one being compiled is unreasonably expensive, so `reifyInstances` will report only the instance for modules that GHC has had some cause to visit during this compilation. This is a shortcoming: `reifyInstances` might fail to report instances for a type that is otherwise unusued, or instances defined in a different component. You can work around this shortcoming by explicitly importing the modules whose instances you want to be visible. GHC issue #20529 has some discussion around this. Fixes #20529
-
