GitLab

Test Plan: validate

Reviewers: austin, bgamari, simonmar, mpickering

Reviewed By: mpickering

Subscribers: mpickering, rwbarton, thomie

GHC Trac Issues: #12610

Differential Revision: https://phabricator.haskell.org/D3584

This reverts commit 7fea7121.
It turns out that on a newly added MultiLayerModules test
case it gets very expensive to union the transitive module
sets while preserving determinism.

Fortunately, we can just sort to restore determinism
when converting imp_dep_mods to a list.

Test Plan: ./validate

Reviewers: simonmar, austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3577

I'm optimizing a case that is well approximated by
multiple layers of modules where every module in a layer
imports all the modules in the layer below.

It turns out I regressed performance on such cases in 7fea7121.
I'm adding a test case to track improvements and prevent
future regressions.

Test Plan: ./validate

Reviewers: simonmar, austin, bgamari

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3575

This reverts commit e527fc2e.

I merged more than I intended in e527fc2e, I will merge
D3575 and D3577 separately.

I'm optimizing a case that is well approximated by
multiple layers of modules where every module in a layer
imports all the modules in the layer below.

It turns out I regressed performance on such cases in 7fea7121.
I'm adding a test case to track improvements and prevent
future regressions.

Test Plan: ./validate

Reviewers: simonmar, austin, bgamari

Reviewed By: simonmar

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3575

GHC build fails for powerpc-unknown-linux-gnu
and hppa-unknown-linux-gnu targets as:
  rts_dist_HC rts/dist/build/RtsStartup.o
    rts/linker/ElfTypes.h:23:4: error:
         error: #error "Unsupported arch!"

Before the change code tried to whitelist architectures
and classify them into ELF32/ELF64. It does not work
for UNREG arches like 'hppa', 'sparc64', 'm68k', 'mips'.

It is nuanced for things like mips64 and x86_64:
    'mips64-unknown-linux-gnu-gcc -mabi=64' is ELFCLASS64
    'mips64-unknown-linux-gnu-gcc'          is ELFCLASS32
    'x86_64-pc-linux-gnu-gcc'               is ELFCLASS64
    'x86_64-pc-linux-gnu-gcc -mx32'         is ELFCLASS32
Here it's not enough to know HOST_ARCH. We really need to
know ABI.

The change uses '__LP64__' as a proxy for ELFCLASS64.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>

Reviewers: angerman, simonmar, austin, bgamari, erikd

Reviewed By: angerman, bgamari, erikd

Subscribers: rwbarton, thomie

GHC Trac Issues: #13696

Differential Revision: https://phabricator.haskell.org/D3583

Fixes gcc-7.1.0 warnings of form:

  rts/sm/Scav.c:559:9: error:
     error: this statement may fall through [-Werror=implicit-fallthrough=]
             scavenge_fun_srt(info);
             ^~~~~~~~~~~~~~~~~~~~~~

Many of places are indeed unobvious and some are
already annotated by comments.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>

Found by gcc-7.1 which reported build error as:

rts/ProfilerReportJson.c:23:16: error:
     error: comparison between pointer and zero
       character constant [-Werror=pointer-compare]
         for (; str != '\0' && len > 0; str++) {
                    ^~
   |
23 |     for (; str != '\0' && len > 0; str++) {
   |                ^

Unfixed code in context:

```c
  static void escapeString(char const* str, char *out, int len)
  {
     len--; // reserve character in output for terminating NUL
     for (; str != '\0' && len > 0; str++) {
         char c = *str;
```

The intent here is to process 'len' (if positive) or '\0'-terminator
in 'str' but dereference was missing.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>

Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #12600

Differential Revision: https://phabricator.haskell.org/D3580

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #11272

Differential Revision: https://phabricator.haskell.org/D3561

This partially reverts commit 37299536 as it
doesn't actually fix #13594. Namely it does not revert the mkPrefixFunRhs
refactoring since this is rather independent from the functional changes.

Going to try again with a whole working patch

Fixes lazy IO bug

Test Plan: validate

Reviewers: austin, bgamari, rwbarton

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13664

Differential Revision: https://phabricator.haskell.org/D3549

Rename isModuleInterpreted to moduleIsBootOrNotObjectLinkable
because a) there already is a moduleIsInterpreted function in
the same module b) I have no idea if the (new) semantic of
the bool returned matches some understanding of
"is interpreted".

Instead of excluding `isAnyInlinePragma`, just exclude
`isInlinePragma`. This makes GHC behave as documented;
the user's guide only indicates that GHC does not automatically
add SCCs to `INLINE` bindings.

Fixes #12962.

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: DemiMarie, osa1, Mikolaj, simonpj, rwbarton, thomie

GHC Trac Issues: #12962

Differential Revision: https://phabricator.haskell.org/D3550

This issue was only ever present in the GHC 8.0.1 release candidates, but
let's add a regression test for it just to be safe.

Commit a7ee2d4c fixed #11966. Here's a
regression test for it.

Summary:
Commit 0c9d9dec (the fix for #13271) fixed the
programs in #12083. This adds regression tests for them.

Test Plan: make test TEST="T12083a T12083b"

Reviewers: austin, bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #12083

Differential Revision: https://phabricator.haskell.org/D3573

Summary: Lots of refactoring in RnEnv to reduce code duplication.

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13545

Differential Revision: https://phabricator.haskell.org/D3507

This adds the GHCi variant of the failing program in #13594. Also, I
inadvertently changed the T13594 test previously introduced in a way that made
it no longer faithfully test the ticket as written. Fix this.

This requirement was introduced around 7.8 but was never documented.

Resolves #9287.

This ended up being quite simple.

Reviewers: austin, goldfire, mpickering

Subscribers: rwbarton, shlevy, thomie

GHC Trac Issues: #13688

Differential Revision: https://phabricator.haskell.org/D3571

`collect_lpat` was missing a case for `HsSplicedPat`, which caused
incorrect renaming of TH-spliced pattern variables.

Fixes #13473.

Test Plan: make test TEST=T13473

Reviewers: facundominguez, austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13473

Differential Revision: https://phabricator.haskell.org/D3572

Fixes #13677 by parenthesizing equalities in a sufficiently high
pretty-printing context.

Test Plan: make test TEST=T13677

Reviewers: goldfire, austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13677

Differential Revision: https://phabricator.haskell.org/D3570

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3565

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3564

StgCRunAsm.S provides StgCRun on powerpc64le platforms when
registerised.  However, in the unregisterised setting we use the
mini-interpreter and consequently shouldn't build StgCRunAsm.S lest we
get duplicate symbols.

Test Plan: Build unregisterised compiler on AIX.

Reviewers: hvr, trommler, austin, simonmar

Reviewed By: trommler, simonmar

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3560

Adds a check in `rnStmt`, in sub-expr `getFailFunction`, to determine if
the pattern of a bind statement is irrefutible. If so, skip looking up
the `fail` name.

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13649

Differential Revision: https://phabricator.haskell.org/D3553

Given multiple in-scope constructors with the same name, say `A`, and a
function of type `A -> Int`, say, the compiler reports both a "type `A`
is not in scope" and (incorrectly) an ambiguity error. The latter
shouldn't be there if `DataKinds` isn't enabled.

This issue was recommended to me by @mpickering as a suitable first
task, and the fix was also outlined in the original Trac ticket. It
involved a simple reordering of the steps taken in `lookup_demoted` in
`RnEnv.hs`. The fix is to make the `DataKinds` check happen earlier,
ensuring that the ambiguity check doesn't happen at all if we know the
constructors couldn't have been promoted.
Signed-off-by: Soham Chowdhury <chow.soham@gmail.com>

Reviewers: mpickering, austin, bgamari

Reviewed By: mpickering, bgamari

Subscribers: rwbarton, thomie

GHC Trac Issues: #13568

Differential Revision: https://phabricator.haskell.org/D3547

Doug Wilson astutely noticed that the pattern match checker was spending
a significant amount of time in mkPmId. It turns out that it was
producing Ids with OccNames that were generated from a Unique, which are
non-trivial to render. Since Var is strict in Name, Name in OccName, and
OccName in its FastString all of this encoding work was being done
despite the fact that it was (as far as I can tell) never actually
needed.

Test Plan: Validate, note allocations of `T11195`

Reviewers: austin, gkaracha

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3568

Commit e7985ed2 happened to fix #12850, so
let's add a regression test for the program reported in #12850.

This reverts commit 8d4bce42.

we do the same for the rts already. And using the configure script should
be more robust than hand-picking the OSs here.

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D3563