[Git][ghc/ghc][wip/nested-cpr-2019] 54 commits: Switch order on `GhcMake.IsBoot`

Sebastian Graf gitlab at gitlab.haskell.org
Mon May 4 16:24:57 UTC 2020 


Sebastian Graf pushed to branch wip/nested-cpr-2019 at Glasgow Haskell Compiler / GHC


Commits:
b0fbfc75 by John Ericson at 2020-04-24T12:07:14-04:00
Switch order on `GhcMake.IsBoot`

In !1798 we were requested to replace many `Bool`s with this data type.
But those bools had `False` meaning `NotBoot`, so the `Ord` instance
would be flipped if we use this data-type as-is.

Since the planned formally-`Bool` occurrences vastly outnumber the
current occurrences, we figured it would be better to conform the `Ord`
instance to how the `Bool` is used now, fixing any issues, rather than
fix them currently with the bigger refactor later in !1798. That way,
!1798 can be a "pure" refactor with no behavioral changes.

- - - - -
af332442 by Sylvain Henry at 2020-04-26T13:55:14-04:00
Modules: Utils and Data (#13009)

Update Haddock submodule

Metric Increase:
   haddock.compiler

- - - - -
cd4434c8 by Sylvain Henry at 2020-04-26T13:55:16-04:00
Fix misleading Ptr phantom type in SerializedCompact (#15653)

- - - - -
22bf5c73 by Ömer Sinan Ağacan at 2020-04-26T13:55:22-04:00
Tweak includes in non-moving GC headers

We don't use hash tables in non-moving GC so remove the includes.

This breaks Compact.c as existing includes no longer include Hash.h, so
include Hash.h explicitly in Compact.c.

- - - - -
99823ed2 by Sylvain Henry at 2020-04-27T20:24:46-04:00
TH: fix Show/Eq/Ord instances for Bytes (#16457)

We shouldn't compare pointer values but the actual bytes.

- - - - -
c62271a2 by Alp Mestanogullari at 2020-04-27T20:25:33-04:00
hadrian: always capture both stdout and stderr when running a builder fails

The idea being that when a builder('s command) fails, we quite likely want to
have all the information available to figure out why. Depending on the builder
_and_ the particular problem, the useful bits of information can be printed
on stdout or stderr.

We accomplish this by defining a simple wrapper for Shake's `cmd` function,
that just _always_ captures both streams in case the command returns a non-zero
exit code, and by using this wrapper everywhere in `hadrian/src/Builder.hs`.

Fixes #18089.

- - - - -
4b9764db by Ryan Scott at 2020-04-28T15:40:04-04:00
Define a Quote IO instance

Fixes #18103.

- - - - -
518a63d4 by Ryan Scott at 2020-04-28T15:40:42-04:00
Make boxed 1-tuples have known keys

Unlike other tuples, which use special syntax and are "known" by way
of a special `isBuiltInOcc_maybe` code path, boxed 1-tuples do not
use special syntax. Therefore, in order to make sure that the
internals of GHC are aware of the `data Unit a = Unit a` definition
in `GHC.Tuple`, we give `Unit` known keys. For the full details, see
`Note [One-tuples] (Wrinkle: Make boxed one-tuple names have known keys)`
in `GHC.Builtin.Types`.

Fixes #18097.

- - - - -
2cfc4ab9 by Sylvain Henry at 2020-04-30T01:56:56-04:00
Document backpack fields in DynFlags

- - - - -
10a2ba90 by Sylvain Henry at 2020-04-30T01:56:56-04:00
Refactor UnitInfo

* Rename InstalledPackageInfo into GenericUnitInfo

The name InstalledPackageInfo is only kept for alleged backward
compatibility reason in Cabal. ghc-boot has its own stripped down copy
of this datatype but it doesn't need to keep the name. Internally we
already use type aliases (UnitInfo in GHC, PackageCacheFormat in
ghc-pkg).

* Rename UnitInfo fields: add "unit" prefix and fix misleading names

* Add comments on every UnitInfo field

* Rename SourcePackageId into PackageId

"Package" already indicates that it's a "source package". Installed
package components are called units.

Update Haddock submodule

- - - - -
69562e34 by Sylvain Henry at 2020-04-30T01:56:56-04:00
Remove unused `emptyGenericUnitInfo`

- - - - -
9e2c8e0e by Sylvain Henry at 2020-04-30T01:56:56-04:00
Refactor UnitInfo load/store from databases

Converting between UnitInfo stored in package databases and UnitInfo as
they are used in ghc-pkg and ghc was done in a very convoluted way (via
BinaryStringRep and DbUnitModuleRep type classes using fun deps, etc.).
It was difficult to understand and even more to modify (I wanted to
try to use a GADT for UnitId but fun deps got in the way).

The new code uses much more straightforward functions to convert between
the different representations. Much simpler.

- - - - -
ea717aa4 by Sylvain Henry at 2020-04-30T01:56:56-04:00
Factorize mungePackagePaths code

This patch factorizes the duplicated code used in ghc-pkg and in GHC to
munge package paths/urls.

It also fixes haddock-html munging in GHC (allowed to be either a file
or a url) to mimic ghc-pkg behavior.

- - - - -
10d15f1e by Sylvain Henry at 2020-04-30T01:56:56-04:00
Refactoring unit management code

Over the years the unit management code has been modified a lot to keep
up with changes in Cabal (e.g. support for several library components in
the same package), to integrate BackPack, etc. I found it very hard to
understand as the terminology wasn't consistent, was referring to past
concepts, etc.

The terminology is now explained as clearly as I could in the Note
"About Units" and the code is refactored to reflect it.

-------------------

Many names were misleading: UnitId is not an Id but could be a virtual
unit (an indefinite one instantiated on the fly), IndefUnitId
constructor may contain a definite instantiated unit, etc.

   * Rename IndefUnitId into InstantiatedUnit
   * Rename IndefModule into InstantiatedModule
   * Rename UnitId type into Unit
   * Rename IndefiniteUnitId constructor into VirtUnit
   * Rename DefiniteUnitId constructor into RealUnit
   * Rename packageConfigId into mkUnit
   * Rename getPackageDetails into unsafeGetUnitInfo
   * Rename InstalledUnitId into UnitId

Remove references to misleading ComponentId: a ComponentId is just an
indefinite unit-id to be instantiated.

   * Rename ComponentId into IndefUnitId
   * Rename ComponentDetails into UnitPprInfo
   * Fix display of UnitPprInfo with empty version: this is now used for
     units dynamically generated by BackPack

Generalize several types (Module, Unit, etc.) so that they can be used
with different unit identifier types: UnitKey, UnitId, Unit, etc.

   * GenModule: Module, InstantiatedModule and InstalledModule are now
     instances of this type
   * Generalize DefUnitId, IndefUnitId, Unit, InstantiatedUnit,
     PackageDatabase

Replace BackPack fake "hole" UnitId by a proper HoleUnit constructor.

Add basic support for UnitKey. They should be used more in the future to
avoid mixing them up with UnitId as we do now.

Add many comments.

Update Haddock submodule

- - - - -
8bfb0219 by Sylvain Henry at 2020-04-30T01:56:56-04:00
Unit: split and rename modules

Introduce GHC.Unit.* hierarchy for everything concerning units, packages
and modules.

Update Haddock submodule

- - - - -
71484b09 by Alexis King at 2020-04-30T01:57:35-04:00
Allow block arguments in arrow control operators

Arrow control operators have their own entries in the grammar, so they
did not cooperate with BlockArguments. This was just a minor oversight,
so this patch adjusts the grammar to add the desired behavior.

fixes #18050

- - - - -
a48cd2a0 by Alexis King at 2020-04-30T01:57:35-04:00
Allow LambdaCase to be used as a command in proc notation

- - - - -
f4d3773c by Alexis King at 2020-04-30T01:57:35-04:00
Document BlockArguments/LambdaCase support in arrow notation

- - - - -
5bdfdd13 by Simon Peyton Jones at 2020-04-30T01:58:15-04:00
Add tests for #17873

- - - - -
19b701c2 by Simon Peyton Jones at 2020-04-30T07:30:13-04:00
Mark rule args as non-tail-called

This was just an omission...b I'd failed to call markAllNonTailCall on
rule args.  I think this bug has been here a long time, but it's quite
hard to trigger.

Fixes #18098

- - - - -
014ef4a3 by Matthew Pickering at 2020-04-30T07:30:50-04:00
Hadrian: Improve tool-args command to support more components

There is a new command to hadrian, tool:path/to/file.hs, which returns
the options needed to compile that file in GHCi.

This is now used in the ghci script with argument `ghc/Main.hs` but its
main purpose is to support the new multi-component branch of ghcide.

- - - - -
2aa67611 by Ben Gamari at 2020-04-30T21:34:44-04:00
nonmoving: Clear bitmap after initializing block size

Previously nonmovingInitSegment would clear the bitmap before
initializing the segment's block size. This is broken since
nonmovingClearBitmap looks at the segment's block size to determine how
much bitmap to clear.

- - - - -
54dad3cf by Ben Gamari at 2020-04-30T21:34:44-04:00
nonmoving: Explicitly memoize block count

A profile cast doubt on whether the compiler hoisted the bound out the
loop as I would have expected here. It turns out it did but nevertheless
it seems clearer to just do this manually.

- - - - -
99ff8145 by Ben Gamari at 2020-04-30T21:34:44-04:00
nonmoving: Eagerly flush all capabilities' update remembered sets

(cherry picked from commit 2fa79119570b358a4db61446396889b8260d7957)

- - - - -
05b0a9fd by Ömer Sinan Ağacan at 2020-04-30T21:35:24-04:00
Remove OneShotInfo field of LFReEntrant, document OneShotInfo

The field is only used in withNewTickyCounterFun and it's easier to
directly pass a parameter for one-shot info to withNewTickyCounterFun
instead of passing it via LFReEntrant. This also makes !2842 simpler.

Other changes:

- New Note (by SPJ) [OneShotInfo overview] added.
- Arity argument of thunkCode removed as it's always 0.

- - - - -
a43620c6 by Ömer Sinan Ağacan at 2020-04-30T21:35:24-04:00
GHC.StgToCmm.Ticky: remove a few unused stuff

- - - - -
780de9e1 by Sylvain Henry at 2020-05-01T10:37:39-04:00
Use platform in Iface Binary

- - - - -
f8386c7b by Sylvain Henry at 2020-05-01T10:37:39-04:00
Refactor PprDebug handling

If `-dppr-debug` is set, then PprUser and PprDump styles are silently
replaced with PprDebug style. This was done in `mkUserStyle` and
`mkDumpStyle` smart constructors. As a consequence they needed a
DynFlags parameter.

Now we keep the original PprUser and PprDump styles until they are used
to create an `SDocContext`. I.e. the substitution is only performed in
`initSDocContext`.

- - - - -
b3df9e78 by Sylvain Henry at 2020-05-01T10:37:39-04:00
Remove PprStyle param of logging actions

Use `withPprStyle` instead to apply a specific style to a SDoc.

- - - - -
de9fc995 by Sylvain Henry at 2020-05-01T10:37:39-04:00
Fully remove PprDebug

PprDebug was a pain to deal with consistently as it is implied by
`-dppr-debug` but it isn't really a PprStyle. We remove it completely
and query the appropriate SDoc flag instead (`sdocPprDebug`) via
helpers (`getPprDebug` and its friends).

- - - - -
8b51fcbd by Sebastian Graf at 2020-05-01T10:38:16-04:00
PmCheck: Only call checkSingle if we would report warnings

- - - - -
fd7ea0fe by Sebastian Graf at 2020-05-01T10:38:16-04:00
PmCheck: Pick up `EvVar`s bound in `HsWrapper`s for long-distance info

`HsWrapper`s introduce evidence bindings through `WpEvLam` which the
pattern-match coverage checker should be made aware of.

Failing to do so caused #18049, where the resulting impreciseness of
imcompleteness warnings seemingly contradicted with
`-Winaccessible-code`.

The solution is simple: Collect all the evidence binders of an
`HsWrapper` and add it to the ambient `Deltas` before desugaring
the wrapped expression.

But that means we pick up many more evidence bindings, even when they
wrap around code without a single pattern match to check! That regressed
`T3064` by over 300%, so now we are adding long-distance info lazily
through judicious use of `unsafeInterleaveIO`.

Fixes #18049.

- - - - -
7bfe9ac5 by Ben Gamari at 2020-05-03T04:41:33-04:00
rts: Enable tracing of nonmoving heap census with -ln

Previously this was not easily available to the user. Fix this.
Non-moving collection lifecycle events are now reported with -lg.

- - - - -
c560dd07 by Ben Gamari at 2020-05-03T04:41:33-04:00
users guide: Move eventlog documentation users guide

- - - - -
02543d5e by Ben Gamari at 2020-05-03T04:41:33-04:00
users guide: Add documentation for non-moving GC events

- - - - -
b465dd45 by Alexis King at 2020-05-03T04:42:12-04:00
Flatten nested casts in the simple optimizer

Normally, we aren’t supposed to generated any nested casts, since mkCast
takes care to flatten them, but the simple optimizer didn’t use mkCast,
so they could show up after inlining. This isn’t really a problem, since
the simplifier will clean them up immediately anyway, but it can clutter
the -ddump-ds output, and it’s an extremely easy fix.

closes #18112

- - - - -
5189d36b by Sebastian Graf at 2020-05-03T22:24:23+02:00
Nested CPR

- - - - -
bd709eed by Sebastian Graf at 2020-05-03T22:24:25+02:00
Move tests from stranal to cpranal

- - - - -
8b10d824 by Sebastian Graf at 2020-05-03T22:24:25+02:00
Accept FacState

- - - - -
ab1d9661 by Sebastian Graf at 2020-05-03T22:26:22+02:00
Factor Cpr and Termination into a joint lattice

As a result, we don't even have to export Termination from Cpr. Neat!

Also I realised there is a simpler and more sound way to generate and
unleash CPR signatures.

- - - - -
32696ae3 by Sebastian Graf at 2020-05-03T22:29:08+02:00
Consider unboxing effects of WW better and get rid of hack

- - - - -
db52466c by Sebastian Graf at 2020-05-03T22:29:10+02:00
stuff

- - - - -
08923728 by Sebastian Graf at 2020-05-03T22:29:10+02:00
Debug output

- - - - -
5fa39431 by Sebastian Graf at 2020-05-03T22:29:11+02:00
A slew of testsuite changes

- - - - -
bab525de by Sebastian Graf at 2020-05-03T22:29:11+02:00
Fix T1600

- - - - -
31f4f11b by Sebastian Graf at 2020-05-03T22:29:11+02:00
Inline `decodeDoubleInteger` and constant-fold `decodeDouble_Int64#` instead

Currently, `decodeDoubleInteger` 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`/`CONSTANT_FOLDED` things since #13143.
Also it is a trade-off: The implementation of `decodeDoubleInteger`
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 `decodeDoubleInteger`.

You may wonder how this affects performance of code using
`integer-simple`; Apparently, according to @hsyl20 this is not a concern
since we will hopefully land !2231 soon.

Fixes #18092.

- - - - -
8427dd1c by Sebastian Graf at 2020-05-03T22:29:11+02:00
Fix primop termination

- - - - -
26599e89 by Sebastian Graf at 2020-05-03T22:29:11+02:00
Test for DataCon wrapper CPR

- - - - -
f1cebd39 by Sebastian Graf at 2020-05-03T22:29:11+02:00
Fix CPR of bottoming functions/primops

- - - - -
db9873c0 by Sebastian Graf at 2020-05-03T22:29:11+02:00
Fix DataConWrapperCpr and accept other test outputs

- - - - -
9e20dd85 by Sebastian Graf at 2020-05-03T22:29:11+02:00
Accept two more changed test outputs

- - - - -
6d523cc8 by Sebastian Graf at 2020-05-03T22:29:11+02:00
Update CaseBinderCPR with a new function

- - - - -
910edd76 by Sebastian Graf at 2020-05-03T22:29:38+02:00
Don't give the case binder the CPR property

- - - - -
91b545f5 by Sebastian Graf at 2020-05-04T18:24:42+02:00
Limit CPR on DataCon app bindings

- - - - -


30 changed files:

- compiler/GHC.hs
- compiler/GHC/Builtin/Names.hs
- compiler/GHC/Builtin/Names.hs-boot
- compiler/GHC/Builtin/Names/TH.hs
- compiler/GHC/Builtin/PrimOps.hs
- compiler/GHC/Builtin/PrimOps.hs-boot
- compiler/GHC/Builtin/Types.hs
- compiler/GHC/Builtin/Types/Literals.hs
- compiler/GHC/Builtin/Types/Prim.hs
- compiler/GHC/Builtin/Uniques.hs
- compiler/GHC/Builtin/Uniques.hs-boot
- compiler/GHC/Builtin/Utils.hs
- compiler/GHC/ByteCode/Asm.hs
- compiler/GHC/ByteCode/InfoTable.hs
- compiler/GHC/ByteCode/Instr.hs
- compiler/GHC/ByteCode/Linker.hs
- compiler/GHC/ByteCode/Types.hs
- compiler/GHC/Cmm.hs
- compiler/GHC/Cmm/BlockId.hs
- compiler/GHC/Cmm/CLabel.hs
- compiler/GHC/Cmm/CallConv.hs
- compiler/GHC/Cmm/CommonBlockElim.hs
- compiler/GHC/Cmm/ContFlowOpt.hs
- compiler/GHC/Cmm/Dataflow.hs
- compiler/GHC/Cmm/Dataflow/Block.hs
- compiler/GHC/Cmm/Dataflow/Collections.hs
- compiler/GHC/Cmm/Dataflow/Graph.hs
- compiler/GHC/Cmm/Dataflow/Label.hs
- compiler/GHC/Cmm/DebugBlock.hs
- compiler/GHC/Cmm/Expr.hs


The diff was not included because it is too large.


View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e64516e7a710c7402660b9a05bdf97fa8f7d5134...91b545f5e2aefede98603d65b2a7056558d27880

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e64516e7a710c7402660b9a05bdf97fa8f7d5134...91b545f5e2aefede98603d65b2a7056558d27880
You're receiving this email because of your account on gitlab.haskell.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20200504/fc7726e3/attachment-0001.html>

More information about the ghc-commits mailing list