Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

This corrects the `jump islands` calculations for Windows.  The code was
incorrectly creating a new entry for every `usage` of a symbol instead
of every used symbol. e.g. if a symbol is used 5 times it used to create
5 jump islands. This is incorrect and not in line with what the `ELF`
and `Mach-O` linkers do. Also since we allocate `n` spaces where `n` is
number of symbols, we would quickly run out of space and abort.

Test Plan: ./validate

Reviewers: simonmar, hvr, erikd, bgamari, austin

Reviewed By: bgamari

Subscribers: thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D3026

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

When the `GCC` driver envokes the pipeline a `SPEC` is used to determine
how to configure the compiler and which libraries to pass along.

For Windows/mingw, this specfile is
https://github.com/gcc-mirror/gcc/blob/master/gcc/config/i386/mingw32.h

This expands the list of base DLLs with the ones that GCC always links,
and adds extra sibling dlls of `stdc++` in case it is linked in.

Following D3028 this patch only needs to load the always load only the
top level individual shared libs.

Test Plan: ./validate

Reviewers: RyanGlScott, austin, bgamari, erikd, simonmar

Reviewed By: bgamari

Subscribers: RyanGlScott, thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D3029

Siddharth authored 8 years ago and Ben Gamari committed 8 years ago

- Fixes (#12636).
- changes all the typecasts to _unsinged long long_ to
  have the format specifiers work.

Reviewers: austin, bgamari, erikd, simonmar, Phyx

Reviewed By: erikd, Phyx

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3129

olsner authored 8 years ago and Ben Gamari committed 8 years ago

Previously this was added only to the RTS's C files (those are the bulk
of it though), but there are C bits in ghc-prim, integer-gmp and base
too.

Followup for #8405, allows the large table of character properties in
base to be stripped when not used.

Test Plan: validate

Reviewers: austin, bgamari, simonmar

Reviewed By: bgamari

Subscribers: thomie, snowleopard

Differential Revision: https://phabricator.haskell.org/D3121

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

Here we move the actual report generation logic to
`rts/ProfilerReport.c`. This break is actually quite clean,

    void writeCCSReport( FILE *prof_file, CostCentreStack const *ccs,
                         ProfilerTotals totals );

This is more profiler refactoring in preparation for machine-readable
output.

Test Plan: Validate

Reviewers: austin, erikd, simonmar

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3097

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

Previously it was quite difficult to follow the dataflow through this
file due to global mutation and rather non-descriptive types.

This is a cleanup in preparation for factoring out the report-generating
logic, which is itself in preparation for somedayteaching the profiler
to produce more machine-readable reports (JSON perhaps?).

Test Plan: Validate

Reviewers: austin, erikd, simonmar

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3096

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

This corrects the unwind information for `stg_stop_thread`, which
allows us to unwind back to the C stack after reaching the end of the 
STG stack.

Test Plan: Validate

Reviewers: simonmar, austin, erikd

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2746

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

And use to mark `stg_stack_underflow_frame`, which we are unable to
determine a caller from.

To simplify parsing at the moment we steal the `return` keyword to
indicate an undefined unwind value. Perhaps this should be revisited.

Reviewers: scpmw, simonmar, austin, erikd

Subscribers: dfeuer, thomie

Differential Revision: https://phabricator.haskell.org/D2738

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

As discussed in D1532, Trac Trac #11337, and Trac Trac #11338, the stack
unwinding information produced by GHC is currently quite approximate.
Essentially we assume that register values do not change at all within a
basic block. While this is somewhat true in normal Haskell code, blocks
containing foreign calls often break this assumption. This results in
unreliable call stacks, especially in the code containing foreign calls.
This is worse than it sounds as unreliable unwinding information can at
times result in segmentation faults.

This patch set attempts to improve this situation by tracking unwinding
information with finer granularity. By dispensing with the assumption of
one unwinding table per block, we allow the compiler to accurately
represent the areas surrounding foreign calls.

Towards this end we generalize the representation of unwind information
in the backend in three ways,

 * Multiple CmmUnwind nodes can occur per block

 * CmmUnwind nodes can now carry unwind information for multiple
   registers (while not strictly necessary; this makes emitting
   unwinding information a bit more convenient in the compiler)

 * The NCG backend is given an opportunity to modify the unwinding
   records since it may need to make adjustments due to, for instance,
   native calling convention requirements for foreign calls (see
   #11353).

This sets the stage for resolving #11337 and #11338.

Test Plan: Validate

Reviewers: scpmw, simonmar, austin, erikd

Subscribers: qnikst, thomie

Differential Revision: https://phabricator.haskell.org/D2741

Takenobu Tani authored 8 years ago and Matthew Pickering committed 8 years ago

[skip ci]

There ware some old file names (.lhs, ...) at comments.

* rts/win32/ThrIOManager.c
  - Conc.lhs -> Conc.hs

* rts/PrimOps.cmm
  - ByteCodeLink.lhs -> ByteCodeLink.hs
  - StgMiscClosures.hc -> StgMiscClosures.cmm

* rts/AutoApply.h
  - AutoApply.hc -> AutoApply.cmm

* rts/HeapStackCheck.cmm
  - PrimOps.hc -> PrimOps.cmm

* rts/LdvProfile.h
  - Updates.hc -> Updates.cmm

* rts/Schedule.c
  - StgStartup.hc -> StgStartup.cmm

* rts/Weak.c
  - StgMiscClosures.hc -> StgMiscClosures.cmm

Reviewers: bgamari, austin, erikd, simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3075

This patch converts the 4 lasting static flags (read from the command
line and unsafely stored in immutable global variables) into dynamic
flags. Most use cases have been converted into reading them from a DynFlags.

In cases for which we don't have easy access to a DynFlags, we read from
'unsafeGlobalDynFlags' that is set at the beginning of each 'runGhc'.
It's not perfect (not thread-safe) but it is still better as we can
set/unset these 4 flags before each run when using GHC API.

Updates haddock submodule.

Rebased and finished by: bgamari

Test Plan: validate

Reviewers: goldfire, erikd, hvr, austin, simonmar, bgamari

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2839

GHC Trac Issues: #8440

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

Here we add support to GHCi for StaticPointers. This process begins by
adding remote GHCi messages for adding entries to the static pointer
table. We then collect binders needing SPT entries after linking and
send the interpreter a message adding entries with the appropriate
fingerprints.

Test Plan: `make test TEST=StaticPtr`

Reviewers: facundominguez, mboes, simonpj, simonmar, goldfire, austin,
hvr, erikd

Reviewed By: simonpj, simonmar

Subscribers: RyanGlScott, simonpj, thomie

Differential Revision: https://phabricator.haskell.org/D2504

GHC Trac Issues: #12356

Currently eventlog data is always written to a file `progname.eventlog`.
This patch introduces the `flushEventLog` field in `RtsConfig` which
allows to customize the writing of eventlog data.

One possible scenario is the ongoing live-profile-monitor effort by
@NCrashed which slurps all eventlog data through `fluchEventLog`.

`flushEventLog` takes a buffer with eventlog data and its size and
returns `false` (0) in case eventlog data could not be procesed.

Reviewers: simonmar, austin, erikd, bgamari

Reviewed By: simonmar, bgamari

Subscribers: qnikst, thomie, NCrashed

Differential Revision: https://phabricator.haskell.org/D2934

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

The symbol not found error that is triggered
during lazy-loading was a bit chaotic before.

This reformats it a bit to:

```
ghc-stage2.exe:  | E:\...\libLLVMSupport.a: unknown symbol `_ZN4llvm5APIntC1Ejyb'
ghc-stage2.exe:  | E:\...\libLLVMCore.a: unknown symbol `_ZN4llvm5APInt14AssignSlowCaseERKS0_'
ghc-stage2.exe:  | E:\...\libLLVMCore.a: unknown symbol `_ZN4llvm13ConstantRangeC1ENS_5APIntES1_'
ghc-stage2.exe:  | E:\...\libLLVMCore.a: unknown symbol `_ZN4llvm14FoldingSetImplC2Ej'
ghc-stage2.exe:  | E:\...\libLLVMCore.a: unknown symbol `_ZN4llvm15LLVMContextImplD1Ev'
ghc-stage2.exe:  | E:\...\libLLVMLTO.a: unknown symbol `_ZN4llvm11LLVMContextD1Ev'
ghc-stage2.exe:  | E:\...\libLLVMCore.a: unknown symbol `_ZNK4llvm5Value10getContextEv'
ghc-stage2.exe: ^^ Could not load 'LLVMIsMultithreaded', dependency unresolved.
See top entry above.
```

I have also thought about also showing the demangled names, as it may
be useful for the end user.

`libgcc` seems to provide a method for this so we wouldn't need any
extra dependency.

Any thoughts on this or would it not be useful?

Reviewers: austin, erikd, simonmar, bgamari

Reviewed By: bgamari

Subscribers: RyanGlScott, thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D3027

GHC Trac Issues: #13093, #13113

Simon Marlow authored 8 years ago and Ben Gamari committed 8 years ago

This isn't supported, and fatalling with an error is better than
segfaulting later.

Test Plan: validate

Reviewers: JonCoens, austin, erikd, niteria, bgamari

Reviewed By: niteria, bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3020

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

Clean up the linker code for PE.

1. Stop copying structures from the windows header
   and use those that are in the headers. There's no
   point in copying them and we got a few types wrong.

2. Replace custom typedef with C99 types. If we're not
   going to use the Windows type aliases, at least use
   standard ones.

Test Plan: ./validate

Reviewers: simonmar, austin, erikd, bgamari

Reviewed By: simonmar, bgamari

Subscribers: dfeuer, thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D2944

Summary:
Someone committed a new public symbol `purgeObj` again
without adding it to the symbols table.

Test Plan: ./validate

Reviewers: austin, bgamari, simonmar, erikd

Reviewed By: erikd

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2980

Summary:
This patch is courtesy of @awson.

Currently, whenever GHC catches a segfault on Windows, it simply reports the
somewhat uninformative message
`Segmentation fault/access violation in generated code`. This patch adds to
the message the type of violation (read/write/dep) and location information,
which should help debugging segfaults in the future.

Fixes #13108.

Test Plan: Build on Windows

Reviewers: austin, erikd, bgamari, simonmar, Phyx

Reviewed By: bgamari, Phyx

Subscribers: awson, thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D2969

GHC Trac Issues: #13108

Summary:
Apparently `sysErrorBelch` doesn't terminate the program anymore making
previously unreachable code now execute. If a dll is not found the error
message we return needs to be a heap value.

Secondly also allow the pattern `lib<name>` to be allowed for finding an
import library with the name `lib<name>.dll.a`.

Test Plan: ./validate, new tests T13082_good and T13082_fail

Reviewers: austin, RyanGlScott, hvr, erikd, simonmar, bgamari

Reviewed By: RyanGlScott, bgamari

Subscribers: thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D2941

GHC Trac Issues: #13082

dobenour authored 8 years ago and Ben Gamari committed 8 years ago

This changes heap overflow to throw a HeapOverflow exception instead of
killing the process.

Test Plan: GHC CI

Reviewers: simonmar, austin, hvr, erikd, bgamari

Reviewed By: simonmar, bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2790

GHC Trac Issues: #1791

* In stg_ap_0_fast, if we're evaluating a thunk, the thunk might
  evaluate to a function in which case we may have to adjust its CCS.

* The interpreter has its own implementation of stg_ap_0_fast, so we
  have to do the same shenanigans with creating empty PAPs and copying
  PAPs there.

* GHCi creates Cost Centres as children of CCS_MAIN, which enterFunCCS()
  wrongly assumed to imply that they were CAFs.  Now we use the is_caf
  flag for this, which we have to correctly initialise when we create a
  Cost Centre in GHCi.

Peter Trommler authored 8 years ago and Ben Gamari committed 8 years ago

The Power ISA says the result of a division by zero is undefined.  So
ignore stdout on PowerPC 64-bit systems.

Disable ext-interp tests on 64-bit PowerPC.  We don't have support for
PowerPC 64-bit ELF in the RTS linker, which is needed for the external
interpreter.

Test Plan: ./validate

Reviewers: austin, simonmar, hvr, erikd, bgamari

Reviewed By: bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2782

gracjan authored 8 years ago and Ben Gamari committed 8 years ago

Reviewers: simonmar, erikd, austin, bgamari

Reviewed By: bgamari

Subscribers: gracjan, rwbarton, thomie

Differential Revision: https://phabricator.haskell.org/D2881

GHC Trac Issues: #13005

Simon Marlow authored 8 years ago and Ben Gamari committed 8 years ago

I forgot to account for BCOs, which have a different layout from
functions.  This caused crashes when using profiling with GHCi (via
-fexternal-interpreter -prof), which unfortunately is not tested at all
by validate, even when profiling is enabled.  I'm going to add some
testing that would have caught this in a separate patch.

Test Plan:
```
cd nofib/spectral/puzzle && make NoFibWithGHCi=YES
EXTRA_RUNTEST_OPTS='-fexternal-interpreter -prof'
```
New testsuite tests coming in a separate diff.

Reviewers: niteria, austin, erikd, bgamari

Reviewed By: bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2868

GHC Trac Issues: #5654

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

This code has been broken on 64-bit systems for some time: the length
and timeout arguments of `addIORequest` and `addDelayRequest`,
respectively, were declared as `int`. However, they were passed Haskell
integers from their respective primops. Integer overflow and madness
ensued. This resulted in #7325 and who knows what else.

Also, there were a few left-over `BOOL`s in here which were not passed
to Windows system calls; these were changed to C99 `bool`s.

However, there is still a bit of signedness inconsistency within the
`delay#` call-chain,

 * `GHC.Conc.IO.threadDelay` and the `delay#` primop accept `Int`
   arguments

 * The `delay#` implementation in `PrimOps.cmm` expects the timeout as
   a `W_`

 * `AsyncIO.c:addDelayRequest` expects an `HsInt` (was `int` prior to
   this patch)

 * `IOManager.c:AddDelayRequest` expects an `HsInt`` (was `int`)

 * The Windows `Sleep` function expects a `DWORD` (which is unsigned)

Test Plan: Validate on Windows

Reviewers: erikd, austin, simonmar, Phyx

Reviewed By: Phyx

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2861

GHC Trac Issues: #7325

The change does the following:
- Add explicit declaration of exception closures
  from base. C backend needs those symbols to be
  visible.
- Reorder cmm functions in use order. Again C
  backend needs symbol declaration/definition
  before use. even for module-local cmm functions.

Fixes the following build failure:

  rts_dist_HC rts/dist/build/Compact.o
    In file included from /tmp/ghc3348_0/ghc_4.hc:3:0: error:
    /tmp/ghc3348_0/ghc_4.hc: In function 'stg_compactAddWithSharingzh':

    /tmp/ghc3348_0/ghc_4.hc:27:11: error:
     error: 'stg_compactAddWorkerzh' undeclared (first use in this function)
     JMP_((W_)&stg_compactAddWorkerzh);
               ^
    ...
    /tmp/ghc3348_0/ghc_4.hc:230:13: error:
     error: 'base_GHCziIOziException_cannotCompactMutable_closure'
     undeclared (first use in this function)
     R1.w = (W_)&base_GHCziIOziException_cannotCompactMutable_closure;
                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Sergei Trofimovich <siarheit@google.com>

Simon Marlow authored 8 years ago and Ben Gamari committed 8 years ago

This fixes some cases of wrong stacks being generated by the profiler.
For background and details on the fix see
`Note [Evaluating functions with profiling]` in `rts/Apply.cmm`.

This does have an impact on allocations for some programs when
profiling.  nofib results:

```
   k-nucleotide          +0.0%     +8.8%    +11.0%    +11.0%      0.0%
         puzzle          +0.0%    +12.5%     0.244     0.246      0.0%
      typecheck           0.0%     +8.7%    +16.1%    +16.2%      0.0%
------------------------------------------------------------------------
--------
            Min          -0.0%     -0.0%    -34.4%    -35.5%    -25.0%
            Max          +0.0%    +12.5%    +48.9%    +49.4%    +10.6%
 Geometric Mean          +0.0%     +0.6%     +2.0%     +1.8%     -0.3%

```

But runtimes don't seem to be affected much, and the examples I looked
at were completely legitimate.  For example, in puzzle we have this:

```
position :: ItemType -> StateType ->  BankType
position Bono = bonoPos
position Edge = edgePos
position Larry = larryPos
position Adam = adamPos
```

where the identifiers on the rhs are all record selectors.  Previously
the profiler gave a stack that looked like

```
  position
  bonoPos
  ...
```

i.e. `bonoPos` was at the same level of the call stack as `position`,
but now it looks like

```
  position
   bonoPos
   ...
```

I used the normaliser from the testsuite to diff the profiling output
from other nofib programs and they all looked better.

Test Plan:
* the broken test passes
* validate
* compiled and ran all of nofib, measured perf, diff'd several .prof
files

Reviewers: niteria, erikd, austin, scpmw, bgamari

Reviewed By: bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2804

GHC Trac Issues: #5654, #10007

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

Mingw-w64 does a stupid thing. They set the FPU precision to extended
mode by default.  The reasoning is that it's for compatibility with GNU
Linux ported libraries. However the problem is this is incompatible with
the standard Windows double precision mode.  In fact, if we create a new
OS thread then Windows will reset the FPU to double precision mode.  So
we end up with a weird state where the main thread by default has a
different precision than any child threads.

Test Plan: ./validate new test T7289

Reviewers: simonmar, austin, bgamari, erikd

Reviewed By: simonmar

Subscribers: thomie, #ghc_windows_task_force

Differential Revision: https://phabricator.haskell.org/D2819

GHC Trac Issues: #7289

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

Test Plan: ./validate

Reviewers: austin, bgamari, erikd, simonmar

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2825

Moritz Angermann authored 8 years ago and Tamar Christina committed 8 years ago

Summary:
The use of globals is quite painful when multiple rts are loaded, e.g.
when plugins are loaded, which bring in a second rts. The sharedCAF
appraoch was employed for the FastStringTable; I've taken the libery
to extend this to the other globals I could find.

This is a reboot of D2575, that should hopefully not exhibit the same
windows build issues.

Reviewers: Phyx, simonmar, goldfire, bgamari, austin, hvr, erikd

Reviewed By: Phyx, simonmar, bgamari

Subscribers: mpickering, thomie

Differential Revision: https://phabricator.haskell.org/D2773

Tamar Christina authored 8 years ago and Ben Gamari committed 8 years ago

Reviewers: bgamari, austin, erikd, simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2818

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

Test Plan: Validate on Windows, ensure that linking works.

Reviewers: Phyx, austin, erikd, simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2817

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

This was removed in 8dc72f3c which cleaned up
PosixSource.h. Strangely, this only started breaking for me now.

Test Plan: Validate on Windows

Reviewers: simonmar, erikd, austin, Phyx

Reviewed By: Phyx

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2812

GHC Trac Issues: #12951

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

Test Plan: Validate on Windows.

Reviewers: austin, erikd, simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2798

GHC Trac Issues: #12388

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

This patch replaces calls to barf() in loadArchive() with proper
error handling.

Test Plan: GHC CI

Reviewers: rwbarton, erikd, hvr, austin, simonmar, bgamari

Reviewed By: bgamari

Subscribers: thomie

Tags: #ghc

Differential Revision: https://phabricator.haskell.org/D2652

GHC Trac Issues: #12388

Ben Gamari authored 8 years ago and Ben Gamari committed 8 years ago

We want to avoid using SIGALRM whenever possible since we will interrupt
long-running system calls. See #10840.

Test Plan: Validate on Darwin

Reviewers: austin, erikd, simonmar

Reviewed By: simonmar

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2796

GHC Trac Issues: #10840