summaryrefslogtreecommitdiff
path: root/compiler/cmm
Commit message (Collapse)AuthorAgeFilesLines
...
* Rearrange error msgs and add section markers (Trac #11014).Evan Laforge2015-11-241-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This puts the "Relevant bindings" section at the end. It uses a TcErrors.Report Monoid to divide messages by importance and then mappends them together. This is not the most efficient way since there are various intermediate Reports and list appends, but it probably doesn't matter since error messages shouldn't get that large, and are usually prepended. In practice, everything is `important` except `relevantBindings`, which is `supplementary`. ErrMsg's errMsgShortDoc and errMsgExtraInfo were extracted into ErrDoc, which has important, context, and suppelementary fields. Each of those three sections is marked with a bullet character, '•' on unicode terminals and '*' on ascii terminals. Since this breaks tons of tests, I also modified testlib.normalise_errmsg to strip out '•'s. --- Additional notes: To avoid prepending * to an empty doc, I needed to filter empty docs. This seemed less error-prone than trying to modify everyone who produces SDoc to instead produce Maybe SDoc. So I added `Outputable.isEmpty`. Unfortunately it needs a DynFlags, which is kind of bogus, but otherwise I think I'd need another Empty case for SDoc, and then it couldn't be a newtype any more. ErrMsg's errMsgShortString is only used by the Show instance, which is in turn only used by Show HscTypes.SourceError, which is in turn only needed for the Exception instance. So it's probably possible to get rid of errMsgShortString, but that would a be an unrelated cleanup. Fixes #11014. Test Plan: see above Reviewers: austin, simonpj, thomie, bgamari Reviewed By: thomie, bgamari Subscribers: simonpj, nomeata, thomie Differential Revision: https://phabricator.haskell.org/D1427 GHC Trac Issues: #11014
* Output source notes in extended DWARF DIEsBen Gamari2015-11-231-0/+6
| | | | | | | | In order to accomplish this we need to ensure that emit DIEs for all DebugBlocks, even those that have been optimized out, lest we end up with undefined symbols of parents at link time. Differential Revision: https://phabricator.haskell.org/D1279
* Dwarf: Ensure tick parentage is preservedBen Gamari2015-11-231-0/+3
| | | | Differential Revision: https://phabricator.haskell.org/D1387
* Support multiple debug output levelsBen Gamari2015-11-231-1/+1
| | | | | | | | | We now only strip block information from DebugBlocks when compiling with `-g1`, intended to be used when only minimal debug information is desired. `-g2` is assumed when `-g` is passed without any integer argument. Differential Revision: https://phabricator.haskell.org/D1281
* Implement function-sections for Haskell code, #8405Simon Brenner2015-11-126-21/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds a flag -split-sections that does similar things to -split-objs, but using sections in single object files instead of relying on the Satanic Splitter and other abominations. This is very similar to the GCC flags -ffunction-sections and -fdata-sections. The --gc-sections linker flag, which allows unused sections to actually be removed, is added to all link commands (if the linker supports it) so that space savings from having base compiled with sections can be realized. Supported both in LLVM and the native code-gen, in theory for all architectures, but really tested on x86 only. In the GHC build, a new SplitSections variable enables -split-sections for relevant parts of the build. Test Plan: validate with both settings of SplitSections Reviewers: dterei, Phyx, austin, simonmar, thomie, bgamari Reviewed By: simonmar, thomie, bgamari Subscribers: hsyl20, erikd, kgardas, thomie Differential Revision: https://phabricator.haskell.org/D1242 GHC Trac Issues: #8405
* Make GHCi & TH work when the compiler is built with -profSimon Marlow2015-11-071-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Amazingly, there were zero changes to the byte code generator and very few changes to the interpreter - mainly because we've used good abstractions that hide the differences between profiling and non-profiling. So that bit was pleasantly straightforward, but there were a pile of other wibbles to get the whole test suite through. Note that a compiler built with -prof is now like one built with -dynamic, in that to use TH you have to build the code the same way. For dynamic, we automatically enable -dynamic-too when TH is required, but we don't have anything equivalent for profiling, so you have to explicitly use -prof when building code that uses TH with a profiled compiler. For this reason Cabal won't work with TH. We don't expect to ship a profiled compiler, so I think that's OK. Test Plan: validate with GhcProfiled=YES in validate.mk Reviewers: goldfire, bgamari, rwbarton, austin, hvr, erikd, ezyang Reviewed By: ezyang Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1407 GHC Trac Issues: #4837, #545
* cmm: Expose machine's stack and return address registerBen Gamari2015-11-013-0/+21
| | | | | | | | | | We will need to use these to setup proper unwinding information for the stg_stop_thread closure. This pokes a hole in the STG abstraction, exposing the machine's stack pointer register so that we can accomplish this. We also expose a dummy return address register, which corresponds to the register used to hold the DWARF return address. Differential Revision: https://phabricator.haskell.org/D1225
* Add subWordC# on x86ishNikita Karetnikov2015-10-312-0/+2
| | | | | | | | | | | | | | | This adds a subWordC# primop which implements subtraction with overflow reporting. Reviewers: tibbe, goldfire, rwbarton, bgamari, austin, hvr Reviewed By: bgamari Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1334 GHC Trac Issues: #10962
* unreg: handle CmmStack in C codegen (unbreaks '-g')Sergei Trofimovich2015-10-301-0/+1
| | | | | | | | The patch makes $ make test TEST="debug T10667" not to fail on CmmStack code generation phase Signed-off-by: Sergei Trofimovich <siarheit@google.com>
* CmmParse: Expose popcnt operationsBen Gamari2015-10-301-1/+6
| | | | | | | | | | | | | | | Make various population count operations available via C-- syntax under the names %popcnt{8,16,32,64}. Fixes #11037. Reviewers: simonmar, austin, ekmett Reviewed By: austin, ekmett Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1402 GHC Trac Issues: #11037
* CmmParse: Clarify description of calling conventionBen Gamari2015-10-171-1/+4
|
* Make Monad/Applicative instances MRP-friendlyHerbert Valerio Riedel2015-10-172-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | This patch refactors pure/(*>) and return/(>>) in MRP-friendly way, i.e. such that the explicit definitions for `return` and `(>>)` match the MRP-style default-implementation, i.e. return = pure and (>>) = (*>) This way, e.g. all `return = pure` definitions can easily be grepped and removed in GHC 8.1; Test Plan: Harbormaster Reviewers: goldfire, alanz, bgamari, quchen, austin Reviewed By: quchen, austin Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1312
* Rename package key to unit ID, and installed package ID to component ID.Edward Z. Yang2015-10-142-26/+26
| | | | | | Comes with Haddock submodule update. Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>
* Debug: Remove extraneous LANGUAGE CPPBen Gamari2015-09-251-1/+1
|
* Annotate CmmBranch with an optional likely targetSimon Marlow2015-09-2311-36/+49
| | | | | | | | | | | | | | | | | Summary: This allows the code generator to give hints to later code generation steps about which branch is most likely to be taken. Right now it is only taken into account in one place: a special case in CmmContFlowOpt that swapped branches over to maximise the chance of fallthrough, which is now disabled when there is a likelihood setting. Test Plan: validate Reviewers: austin, simonpj, bgamari, ezyang, tibbe Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1273
* s/StgArrWords/StgArrBytes/Siddhanathan Shanmugam2015-09-111-2/+2
| | | | | | | | | | Rename StgArrWords to StgArrBytes (see Trac #8552) Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D1233 GHC Trac Issues: #8552
* Refactor: delete most of the module FastTypesThomas Miedema2015-08-211-18/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverses some of the work done in #1405, and goes back to the assumption that the bootstrap compiler understands GHC-haskell. In particular: * use MagicHash instead of _ILIT and _CLIT * pattern matching on I# if possible, instead of using iUnbox unnecessarily * use Int#/Char#/Addr# instead of the following type synonyms: - type FastInt = Int# - type FastChar = Char# - type FastPtr a = Addr# * inline the following functions: - iBox = I# - cBox = C# - fastChr = chr# - fastOrd = ord# - eqFastChar = eqChar# - shiftLFastInt = uncheckedIShiftL# - shiftR_FastInt = uncheckedIShiftRL# - shiftRLFastInt = uncheckedIShiftRL# * delete the following unused functions: - minFastInt - maxFastInt - uncheckedIShiftRA# - castFastPtr - panicDocFastInt and pprPanicFastInt * rename panicFastInt back to panic# These functions remain, since they actually do something: * iUnbox * bitAndFastInt * bitOrFastInt Test Plan: validate Reviewers: austin, bgamari Subscribers: rwbarton Differential Revision: https://phabricator.haskell.org/D1141 GHC Trac Issues: #1405
* Fix unused-matches warnings in CmmLex.xThomas Miedema2015-08-091-4/+4
|
* cmmCreateSwitchPlan: Handle singletons up-frontJoachim Breitner2015-08-081-2/+13
| | | | | | | and make sure these are implemented with an equality check, which is a shorter instruction. This was suggested by rwbarton in #10677. Differential Revision: https://phabricator.haskell.org/D1137
* CmmParse: Don't force alignment in memcpy-ish operationsBen Gamari2015-08-031-2/+1
| | | | | | | | | | | | This was initially made in 681973c31c614185229bdae4f6b7ab4f6e64753d. Here I wanted to enforce that the alignment passed to %memcpy was a constant expression, as this is required by LLVM. However, this breaks the knot-tying done in `loopDecls`, causing T8131 to hang. Here I remove the `seq` and mark T8131 as `expect_broken` in the case of the NCG, which doesn't force the alignment in this case. Fixes #10664.
* Bitmap: Fix thunk explosionBen Gamari2015-07-091-20/+64
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously we would build up another `map (-N)` thunk for every word in the bitmap. Now we strictly accumulate the position and carry out a single ``map (`subtract` accum)``. `Bitmap.intsToBitmap` showed up in the profile while compiling a testcase of #7450 (namely a program containing a record type with large number of fields which derived `Read`). The culprit was `CmmBuildInfoTables.procpointSRT.bitmap`. On the testcase (with 4096 fields), the profile previously looked like, ``` total time = 307.94 secs (307943 ticks @ 1000 us, 1 processor) total alloc = 336,797,868,056 bytes (excludes profiling overheads) COST CENTRE MODULE %time %alloc lintAnnots CoreLint 17.2 25.8 procpointSRT.bitmap CmmBuildInfoTables 11.3 25.2 FloatOutwards SimplCore 7.5 1.6 flatten.lookup CmmBuildInfoTables 4.0 3.9 ... ``` After this fix it looks like, ``` total time = 256.88 secs (256876 ticks @ 1000 us, 1 processor) total alloc = 255,033,667,448 bytes (excludes profiling overheads) COST CENTRE MODULE %time %alloc lintAnnots CoreLint 20.3 34.1 FloatOutwards SimplCore 9.1 2.1 flatten.lookup CmmBuildInfoTables 4.8 5.2 pprNativeCode AsmCodeGen 3.7 4.3 simplLetUnfolding Simplify 3.6 2.2 StgCmm HscMain 3.6 2.1 ``` Signed-off-by: Ben Gamari <ben@smart-cactus.org> Test Plan: Validate Reviewers: austin, simonpj Reviewed By: simonpj Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D1041 GHC Trac Issues: #7450
* Add Note [Warnings in code generated by Alex]Thomas Miedema2015-07-071-5/+1
| | | | Differential Revision: https://phabricator.haskell.org/D1044
* Use -fno-warn-unused-imports instead of hiding `ord`Thomas Miedema2015-07-051-1/+2
| | | | | This is more future proof. Maybe later versions of Alex won't `import Data.Char (ord)` anymore.
* Lexer: remove -fno-warn-unused-do-bindThomas Miedema2015-07-041-6/+3
|
* Replace usages of `-w` by `-fno-warn`sThomas Miedema2015-07-042-12/+7
| | | | | | | | | | | And remove unused imports and language pragmas. I checked that the minimum Happy and Alex version requirements, as listed in aclocal.m4, don't have to change. Before building ghc, I ran: - cabal install happy==1.19.4 --with-ghc=ghc-7.8.4 - cabal install alex==3.1.0 --with-ghc=ghc-7.6.3 Differential Revision: https://phabricator.haskell.org/D1032
* Implement PowerPC 64-bit native code backend for LinuxPeter Trommler2015-07-031-4/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Extend the PowerPC 32-bit native code generator for "64-bit PowerPC ELF Application Binary Interface Supplement 1.9" by Ian Lance Taylor and "Power Architecture 64-Bit ELF V2 ABI Specification -- OpenPOWER ABI for Linux Supplement" by IBM. The latter ABI is mainly used on POWER7/7+ and POWER8 Linux systems running in little-endian mode. The code generator supports both static and dynamic linking. PowerPC 64-bit code for ELF ABI 1.9 and 2 is mostly position independent anyway, and thus so is all the code emitted by the code generator. In other words, -fPIC does not make a difference. rts/stg/SMP.h support is implemented. Following the spirit of the introductory comment in PPC/CodeGen.hs, the rest of the code is a straightforward extension of the 32-bit implementation. Limitations: * Code is generated only in the medium code model, which is also gcc's default * Local symbols are not accessed directly, which seems to also be the case for 32-bit * LLVM does not work, but this does not work on 32-bit either * Must use the system runtime linker in GHCi, because the GHC linker for "static" object files (rts/Linker.c) for PPC 64-bit is not implemented. The system runtime (dynamic) linker works. * The handling of the system stack (register 1) is not ELF- compliant so stack traces break. Instead of allocating a new stack frame, spill code should use the "official" spill area in the current stack frame and deallocation code should restore the back chain * DWARF support is missing Fixes #9863 Test Plan: validate (on powerpc, too) Reviewers: simonmar, trofi, erikd, austin Reviewed By: trofi Subscribers: bgamari, arnons1, kgardas, thomie Differential Revision: https://phabricator.haskell.org/D629 GHC Trac Issues: #9863
* Be aware of overlapping global STG registers in CmmSink (#10521)Reid Barton2015-06-253-20/+68
| | | | | | | | | | | | | | | | | | | Summary: On x86_64, commit e2f6bbd3a27685bc667655fdb093734cb565b4cf assigned the STG registers F1 and D1 the same hardware register (xmm1), and the same for the registers F2 and D2, etc. When mixing calls to functions involving Float#s and Double#s, this can cause wrong Cmm optimizations that assume the F1 and D1 registers are independent. Reviewers: simonpj, austin Reviewed By: austin Subscribers: simonpj, thomie, bgamari Differential Revision: https://phabricator.haskell.org/D993 GHC Trac Issues: #10521
* Encode alignment in MO_Memcpy and friendsBen Gamari2015-06-163-26/+52
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Alignment needs to be a compile-time constant. Previously the code generators had to jump through hoops to ensure this was the case as the alignment was passed as a CmmExpr in the arguments list. Now we take care of this up front. This fixes #8131. Authored-by: Reid Barton <rwbarton@gmail.com> Dusted-off-by: Ben Gamari <ben@smart-cactus.org> Tests for T8131 Test Plan: Validate Reviewers: rwbarton, austin Reviewed By: rwbarton, austin Subscribers: bgamari, carter, thomie Differential Revision: https://phabricator.haskell.org/D624 GHC Trac Issues: #8131
| * Encode alignment in MO_Memcpy and friendsBen Gamari2015-06-163-26/+52
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Alignment needs to be a compile-time constant. Previously the code generators had to jump through hoops to ensure this was the case as the alignment was passed as a CmmExpr in the arguments list. Now we take care of this up front. This fixes #8131. Authored-by: Reid Barton <rwbarton@gmail.com> Dusted-off-by: Ben Gamari <ben@smart-cactus.org> Tests for T8131 Test Plan: Validate Reviewers: rwbarton, austin Reviewed By: rwbarton, austin Subscribers: bgamari, carter, thomie Differential Revision: https://phabricator.haskell.org/D624 GHC Trac Issues: #8131
* | UNREG: fix pprHexVal to emit zeros (#10518)Sergei Trofimovich2015-06-141-2/+3
|/ | | | | | | | | | | | | | | | | | | | | | | | | | | jakzale on #ghc reported a build failure when ported GHC on a new target. The code 'pprHexVal (2^32) W32' emits '0xU' which is invalid C. I've introduced bug in 43f1b2ecd1960fa7377cf55a2b97c66059a701ef when added literal truncation. That truncation is a new source of zeros. Signed-off-by: Sergei Trofimovich <siarheit@google.com> Test Plan: added test and tested on UNREG ghc Reviewers: austin Reviewed By: austin Subscribers: thomie, bgamari Differential Revision: https://phabricator.haskell.org/D987 GHC Trac Issues: #10518
* Squash typos in commentsGabor Greif2015-06-121-1/+1
|
* CmmCommonBlockElim: Improve hash functionJoachim Breitner2015-05-181-45/+26
| | | | | | | | | | | | | | | | Previously, the hash function used to cut down the number of block comparisons did not take local registers into account, causing far too many similar, but different bocks to be considered candidates for the (expensive!) comparision. Adding register to the hash takes CmmCommonBlockElim's share of the runtime of the example in #10397 from 17% to 2.5%, and eliminates all unwanted hash collisions. This patch also replaces the fancy trie by a plain Data.Map. It turned out to be not performance critical, so this simplifies the code. Differential Revision: https://phabricator.haskell.org/D896
* Speed up elimCommonBlocks by grouping blocks also by outgoing labelsJoachim Breitner2015-05-161-31/+112
| | | | | | | | | This is an attempt to improve the situation described in #10397, where the linear scan of possible candidates for commoning up is far too expensive. There is (ever) more room for improvement, but this is a start. Differential Revision: https://phabricator.haskell.org/D892
* Typos in error messages and in commentsGabor Greif2015-04-102-2/+2
|
* CmmSwitch: Do not trip over a case with no (valid) branchesJoachim Breitner2015-04-051-2/+4
| | | | | | This fixes #10245. I did not commit the test case, as it fails unconditionally with a compiler built with -DDEBUG, so maybe it is bogus Haskell anyways.
* Refactor the story around switches (#10137)Joachim Breitner2015-03-3013-75/+591
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This re-implements the code generation for case expressions at the Stg → Cmm level, both for data type cases as well as for integral literal cases. (Cases on float are still treated as before). The goal is to allow for fancier strategies in implementing them, for a cleaner separation of the strategy from the gritty details of Cmm, and to run this later than the Common Block Optimization, allowing for one way to attack #10124. The new module CmmSwitch contains a number of notes explaining this changes. For example, it creates larger consecutive jump tables than the previous code, if possible. nofib shows little significant overall improvement of runtime. The rather large wobbling comes from changes in the code block order (see #8082, not much we can do about it). But the decrease in code size alone makes this worthwhile. ``` Program Size Allocs Runtime Elapsed TotalMem Min -1.8% 0.0% -6.1% -6.1% -2.9% Max -0.7% +0.0% +5.6% +5.7% +7.8% Geometric Mean -1.4% -0.0% -0.3% -0.3% +0.0% ``` Compilation time increases slightly: ``` -1 s.d. ----- -2.0% +1 s.d. ----- +2.5% Average ----- +0.3% ``` The test case T783 regresses a lot, but it is the only one exhibiting any regression. The cause is the changed order of branches in an if-then-else tree, which makes the hoople data flow analysis traverse the blocks in a suboptimal order. Reverting that gets rid of this regression, but has a consistent, if only very small (+0.2%), negative effect on runtime. So I conclude that this test is an extreme outlier and no reason to change the code. Differential Revision: https://phabricator.haskell.org/D720
* Remove comments and flag for GranSimThomas Miedema2015-03-192-10/+0
| | | | | | | | | The GranSim code was removed in dd56e9ab and 297b05a9 in 2009, and perhaps other commits I couldn't find. Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D737
* Remove obsolete comment in cmmOffsetJoachim Breitner2015-03-091-7/+0
| | | | | | | | | | | | | | Summary: as obviously, the code does _not_ adhere to the comment, and yet the compiler does _not_ go into an infinite loop. Test Plan: Run validate Reviewers: austin Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D719
* Replace .lhs with .hs in compiler commentsYuri de Wit2015-02-091-1/+1
| | | | | | | | | | | | | | Summary: It looks like during .lhs -> .hs switch the comments were not updated. So doing exactly that. Reviewers: austin, jstolarek, hvr, goldfire Reviewed By: austin, jstolarek Subscribers: thomie, goldfire Differential Revision: https://phabricator.haskell.org/D621 GHC Trac Issues: #9986
* cmm lex: drop unused 'align' tokenSergei Trofimovich2015-01-202-3/+0
| | | | | | | | | | | | | Max removed 'align' token from parser productions long ago: > commit ec39750268da134c906b8bb4b7d61763d4ea5c2e > Author: Max Bolingbroke <batterseapower@hotmail.com> > Date: Tue Jul 5 09:31:08 2011 +0100 > > Remove the unused CmmAlign and CmmDataLabel from CmmStatic This patch drops 'align' from lexer as well. Signed-off-by: Sergei Trofimovich <siarheit@google.com>
* CMM: add a mechanism to import C .data labelsSergei Trofimovich2015-01-191-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: This introduces new .cmm syntax for import: 'import' 'CLOSURE' <identifier>; Currently cmm syntax allows importing only function labels: import pthread_mutex_lock; but sometimes ghc needs to import global gariables or haskell closures: import ghczmprim_GHCziTypes_True_closure; import base_ControlziExceptionziBase_nestedAtomically_closure; import ghczmprim_GHCziTypes_False_closure; import sm_mutex; It breaks on ia64 where there is a difference in pointers to data and pointer to functions. Patch fixes threaded runtime on ia64 where dereference of 'sm_mutex' from CMM led to incurrect location. Exact breakage machanics are the same as in e18525fae273f4c1ad8d6cbe1dea4fc074cac721 Merge into the 7.10 branch Signed-off-by: Sergei Trofimovich <siarheit@google.com> Test Plan: passes ./validate, makes ghci work on ghc-7.8.4 Reviewers: simonmar, simonpj, austin Reviewed By: austin Subscribers: thomie Differential Revision: https://phabricator.haskell.org/D622
* Remove redundant constraints in the compiler itself, found by ↵Simon Peyton Jones2015-01-062-5/+5
| | | | -fwarn-redundant-constraints
* Fix linker interaction between Template Haskell and HPC (#9762)Reid Barton2014-12-231-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: I'm not really happy about perpetuating the hackish fix for #8696, but at least in the context of building with -fhpc, the performance cost should be negligible. I'm suspicious about PlainModuleInitLabel and the Windows stuff too, but I don't know what it does / can't test it (respectively) so I'll leave those alone for now. Hopefully out-of-process TH will save us from these hacks some day. The test is an adaptation of T8696. It's a bit more awkward since I couldn't think of a way to get cross-module tickbox references without optimizations (inlining), but ghci doesn't permit -O for some reason. Test Plan: harbormaster; validate Reviewers: austin Reviewed By: austin Subscribers: carter, thomie Differential Revision: https://phabricator.haskell.org/D583 GHC Trac Issues: #9762 Conflicts: testsuite/tests/ghci/scripts/all.T
* Eliminate so-called "silent superclass parameters"Simon Peyton Jones2014-12-232-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The purpose of silent superclass parameters was to solve the awkward problem of superclass dictinaries being bound to bottom. See THE PROBLEM in Note [Recursive superclasses] in TcInstDcls Although the silent-superclass idea worked, * It had non-local consequences, and had effects even in Haddock, where we had to discard silent parameters before displaying instance declarations * It had unexpected peformance costs, shown up by Trac #3064 and its test case. In monad-transformer code, when constructing a Monad dictionary you had to pass an Applicative dictionary; and to construct that you neede a Functor dictionary. Yet these extra dictionaries were often never used. (All this got much worse when we added Applicative as a superclass of Monad.) Test T3064 compiled *far* faster after silent superclasses were eliminated. * It introduced new bugs. For example SilentParametersOverlapping, T5051, and T7862, all failed to compile because of instance overlap directly because of the silent-superclass trick. So this patch takes a new approach, which I worked out with Dimitrios in the closing hours before Christmas. It is described in detail in THE PROBLEM in Note [Recursive superclasses] in TcInstDcls. Seems to work great! Quite a bit of knock-on effect * The main implementation work is in tcSuperClasses in TcInstDcls Everything else is fall-out * IdInfo.DFunId no longer needs its n-silent argument * Ditto IDFunId in IfaceSyn * Hence interface file format changes * Now that DFunIds do not have silent superclass parameters, printing out instance declarations is simpler. There is tiny knock-on effect in Haddock, so that submodule is updated * I realised that when computing the "size of a dictionary type" in TcValidity.sizePred, we should be rather conservative about type functions, which can arbitrarily increase the size of a type. Hence the new datatype TypeSize, which has a TSBig constructor for "arbitrarily big". * instDFunType moves from TcSMonad to Inst * Interestingly, CmmNode and CmmExpr both now need a non-silent (Ord r) in a couple of instance declarations. These were previously silent but must now be explicit. * Quite a bit of wibbling in error messages
* Some Dwarf generation fixesPeter Wortmann2014-12-181-4/+7
| | | | | | | | - Make abbrev offset absolute on Non-Mac systems - Add another termination byte at the end of the abbrev section (readelf complains) - Scope combination was wrong for the simpler cases - Shouldn't have a "global/" in front of all scopes
* Generate DWARF info sectionPeter Wortmann2014-12-161-7/+23
| | | | | | | | | | | | | | | | This is where we actually make GHC emit DWARF code. The info section contains all the general meta information bits as well as an entry for every block of native code. Notes: * We need quite a few new labels in order to properly address starts and ends of blocks. * Thanks to Nathan Howell for taking the iniative to get our own Haskell language ID for DWARF! (From Phabricator D396)
* Debug data extraction (NCG support)Peter Wortmann2014-12-161-0/+309
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | The purpose of the Debug module is to collect all required information to generate debug information (DWARF etc.) in the back-ends. Our main data structure is the "debug block", which carries all information we have about a block of code that is going to get produced. Notes: * Debug blocks are arranged into a tree according to tick scopes. This makes it easier to reason about inheritance rules. Note however that tick scopes are not guaranteed to form a tree, which requires us to "copy" ticks to not lose them. * This is also where we decide what source location we regard as representing a code block the "best". The heuristic is basically that we want the most specific source reference that comes from the same file we are currently compiling. This seems to be the most useful choice in my experience. * We are careful to not be too lazy so we don't end up breaking streaming. Debug data will be kept alive until the end of codegen, after all. * We change native assembler dumps to happen right away for every Cmm group. This simplifies the code somewhat and is consistent with how pretty much all of GHC handles dumps with respect to streamed code. (From Phabricator D169)
* Add unwind information to CmmPeter Wortmann2014-12-167-0/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Unwind information allows the debugger to discover more information about a program state, by allowing it to "reconstruct" other states of the program. In practice, this means that we explain to the debugger how to unravel stack frames, which comes down mostly to explaining how to find their Sp and Ip register values. * We declare yet another new constructor for CmmNode - and this time there's actually little choice, as unwind information can and will change mid-block. We don't actually make use of these capabilities, and back-end support would be tricky (generate new labels?), but it feels like the right way to do it. * Even though we only use it for Sp so far, we allow CmmUnwind to specify unwind information for any register. This is pretty cheap and could come in useful in future. * We allow full CmmExpr expressions for specifying unwind values. The advantage here is that we don't have to make up new syntax, and can e.g. use the WDS macro directly. On the other hand, the back-end will now have to simplify the expression until it can sensibly be converted into DWARF byte code - a process which might fail, yielding NCG panics. On the other hand, when you're writing Cmm by hand you really ought to know what you're doing. (From Phabricator D169)
* Tick scopesPeter Wortmann2014-12-1610-72/+209
| | | | | | | | | | | | | | | | | | | | | | This patch solves the scoping problem of CmmTick nodes: If we just put CmmTicks into blocks we have no idea what exactly they are meant to cover. Here we introduce tick scopes, which allow us to create sub-scopes and merged scopes easily. Notes: * Given that the code often passes Cmm around "head-less", we have to make sure that its intended scope does not get lost. To keep the amount of passing-around to a minimum we define a CmmAGraphScoped type synonym here that just bundles the scope with a portion of Cmm to be assembled later. * We introduce new scopes at somewhat random places, aligning with getCode calls. This works surprisingly well, but we might have to add new scopes into the mix later on if we find things too be too coarse-grained. (From Phabricator D169)
* Source notes (Cmm support)Peter Wortmann2014-12-167-14/+98
| | | | | | | | | | | | | | | | | | | | This patch adds CmmTick nodes to Cmm code. This is relatively straight-forward, but also not very useful, as many blocks will simply end up with no annotations whatosever. Notes: * We use this design over, say, putting ticks into the entry node of all blocks, as it seems to work better alongside existing optimisations. Now granted, the reason for this is that currently GHC's main Cmm optimisations seem to mainly reorganize and merge code, so this might change in the future. * We have the Cmm parser generate a few source notes as well. This is relatively easy to do - worst part is that it complicates the CmmParse implementation a bit. (From Phabricator D169)
View Lorry Controller Status