Timeline


and

Mar 28, 2013:

7:31 PM Ticket #7799 (Assembly error while building GHC 7.7) created by Iceland_jack
While building GHC 7.7 (after running `make') I got this error message: …
5:50 PM Ticket #7796 (improve dead code elimination in CorePrep) closed by nfrisby
fixed: I added the test case simplCore/should_compile/T7796 to test for this. …
5:18 PM Changeset in ghc [3333a9c5]ghc-7.8wip/Cabal-1.20wip/T8545-ghc-7.8wip/T8776wip/T8995-level-generalisationwip/nested-cprwip/pattern-synonymswip/recurs-compat by Nicolas Frisby <nicolas.frisby@…>
new test for #7796, also improves a haddock perf T149 is also an unexpected pass, but it's behaving nondeterministically, so I'm not changing it for now
5:18 PM Changeset in testsuite [8697d9a]atomicsdata-kind-syntaxoverlapping-tyfamsth-newtype-nats-simplewip/nested-cprwip/pattern-synonymswip/th-new by Nicolas Frisby <nicolas.frisby@…>
new test for #7796, also improves a haddock perf T149 is also an unexpected pass, but it's behaving nondeterministically, so I'm not changing it for now
9:20 AM Changeset in ghc [c7d80c6]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Nicolas Frisby <nicolas.frisby@…>
improve dead code elimination in CorePrep (fixes #7796)
8:16 AM Ticket #7798 (test_bindist reports ".../libffi-6.dll (No such file or directory)") created by thorkilnaur
Since http://darcs.haskell.org/ghcBuilder/builders/tn23/873.html, the tn23 …
5:00 AM Ticket #7553 (ghc fails to terminate with -O2 or greater) closed by amosrobinson
duplicate: Yes, vector fold with a list accumulator will trigger the ForceSpecConstr?
1:46 AM Changeset in ghc [81d55a9]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Amos Robinson <amos.robinson@…>
Fix non-termination of SpecConstr (see #5550). ForceSpecConstr will now only specialise recursive types a finite number of times. There is a new option -fspec-constr-recursive, with a default value of 3.

Mar 27, 2013:

6:00 PM Changeset in ghc [b7bd3f9]cardinality by Simon Peyton Jones <simonpj@…>
Add new dynamic flags -fkill-absence stops GHC doing absence analysis -fkill-one-shot stops GHC spotting one-shot lambdas and single-entry thunks These are just for performance comparisons
2:24 PM Changeset in ghc [54bb2f8]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Simon Peyton Jones <simonpj@…>
Improve comments about dead code (thanks to Nick Frisby)
2:23 PM Changeset in ghc [273183c6]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Simon Peyton Jones <simonpj@…>
Comment typo
2:23 PM Changeset in ghc [1b37a386]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Simon Peyton Jones <simonpj@…>
Comment only

Mar 26, 2013:

8:59 PM Ticket #7797 (re-enable the defun RULE from a SPECIALISE instance pragma) created by nfrisby
As of commit 51d89a55c3, …
4:14 PM Ticket #7796 (improve dead code elimination in CorePrep) created by nfrisby
#4962 is about generating code for (and allocating at runtime!) bindings …
3:56 PM Ticket #7795 (panic! the impossible happened) closed by monoidal
duplicate: This is caused by deriving (show), which should be deriving (Show). …
2:54 PM Ticket #7795 (panic! the impossible happened) created by guest
Got the following in ghch *Calc> :l Calc [1 of 1] Compiling Calc …
1:13 PM Changeset in ghc [b8b52a15]cardinality by Simon Peyton Jones <simonpj@…>
Make the occurrence analyser propagate info into non-rec RHSs See Note [Use one-shot info] in OccurAnal, second bullet. This one change produced some useful wins Program Size Allocs Runtime Elapsed TotalMem -------------------------------------------------------------------------------- hpg +0.0% -2.4% -4.3% -4.8% +0.0% k-nucleotide -0.0% -7.1% -0.3% -0.3% +0.0% sphere +0.0% -5.6% 0.08 0.08 +0.0% -------------------------------------------------------------------------------- Min -0.1% -7.1% -10.7% -11.0% -33.3% Max +0.1% +0.2% +15.4% +15.4% +1.7% Geometric Mean +0.0% -0.2% -0.1% -0.2% -0.7%
2:26 AM Ticket #7794 (GHCi "Prelude.undefined" exceptions on ARM; ByteCodeItbls.mkJumpToAddr ...) created by cjwatson
A number of non-trivial operations fail in ghci on ARM (specifically armhf …

Mar 25, 2013:

10:12 PM Building/CrossCompiling/iOS edited by StephenBlackheath
Remove adjustor pools documentation because adjustor pools are no more (diff)
1:02 PM Changeset in ghc [e721135]cardinality by Simon Peyton Jones <simonpj@…>
Comments
1:02 PM Changeset in ghc [3c3452ee]cardinality by Simon Peyton Jones <simonpj@…>
Back out of "Note [Used should win]" temporarily
1:01 PM Changeset in ghc [be5fe43]cardinality by Simon Peyton Jones <simonpj@…>
Improve simple-minded boxity analysis for function bodies
12:59 PM Changeset in ghc [3cf6e67d]cardinality by Simon Peyton Jones <simonpj@…>
Fix egregious back-to-front case for argOneShots
11:04 AM Ticket #1518 (Make it possible to evaluate monadic actions when assigning record fields ...) closed by liyang
wontfix: Is this really necessary now that we have RecordWildCards?, NamedFieldPuns?
8:48 AM Building edited by StephenBlackheath
Cross compilation is used both for porting and for cross compilers, so … (diff)
8:17 AM Ticket #7791 (Bad error message when strictness annotations are used without ...) closed by simonpj
invalid: Yes, the trouble is that […] really is a valid definition in Haskell …
8:16 AM CodeOwners edited by simonpj
(diff)
8:02 AM Ticket #7680 (-prof build: trying to 'show' a 'Double' leads to 'Char.intToDigit: not a ...) closed by slyfox
invalid: Rebuilt ghc with gcc-7.4.2-r1 (gentoo's fix for gcc) and can't reproduce …
7:27 AM Changeset in ghc [b9d5373]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Edward Z. Yang <ezyang@…>
Heap profiling: flush .hp file at the end of each frame

Mar 24, 2013:

9:36 PM Ticket #7716 (ZonedTime read instance failing to parse what show returns) closed by igloo
invalid: Thanks for the report. However, please send bugs in the time package to …
9:32 PM Ticket #7731 (Add "source" button for instances in Hackage) closed by igloo
invalid: Thanks for the report. However, haddock issues should be filed in …
9:29 PM Changeset in ghc [98267a86]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Gabor Pali <pali.gabor@…>
Allow to bootstrap with a compiler from the FreeBSD Ports Collection
8:48 PM Ticket #7793 (GHC User’s Guide, Version 7.6.2: Page numbering errors) created by thorkilnaur
In The Glorious Glasgow Haskell Compilation System User’s Guide, Version …
8:32 PM Ticket #7749 (hsc2hs generated executable does not reflect additional library ...) closed by igloo
invalid: This doesn't look like an hsc2hs bug to me. Presumably if you compile a …
8:22 PM Ticket #2555 (Template Haskell does not respect -package and -hide constraints) closed by igloo
worksforme: Agreed; we don't have a way to reproduce it, so closing.
6:55 PM Changeset in ghc [d62b64d]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
missed a bunch of places in my flag changes
5:53 PM Commentary/PlatformNaming edited by mtahmed
(diff)
5:39 PM Changeset in ghc [c445412]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
was not ignoring tyco arguments when determing argument strictness
5:38 PM Changeset in ghc [fc34e657]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
fixed conflict in simpleCore pipeline
5:37 PM Changeset in ghc [604bf0a5]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
added flags for the LLF's use of strictness info and its escape analysis
5:37 PM Changeset in ghc [3a0a6b38]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
added -fno-late-float-abstract-LNE-var flag to disallow abstracting over LNE variables
5:37 PM Changeset in ghc [51b72052]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
now identifying LNEs and not considering them as closures (ie captors)
5:36 PM Changeset in ghc [ace803f]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
preparation in SetLevels for identifying LNEs -- confirmed that it did not change allocation wrt last night's baseline nor cg0pin
5:36 PM Changeset in ghc [1dd6c53]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
cleaned-up flags to stop confusing myself SetLevels now uses strictess information to determine if arguments will become closures, this enables the -8% in cichelli and helps a bit in anna too I think the captors-interpretation plan is very nearly precise except for LNEs now I added the late-float-retry flag, it enables a rough draft in SetLevels that attempts to unpin vacuous cycles involving closure growth and fast call preservation
5:36 PM Changeset in ghc [641ad5d]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
total rewrite of the in-closure analysis: using the interpreted Captors data type for now
5:36 PM Changeset in ghc [aeac1fbc]late-lam-lift by Nicolas Frisby <nicolas.frisby@…>
clean-up in SetLevels, better debug output, tracking the names of thunks that a binding occurs in; next: do not penalize for those thunks if they have been floated!
5:29 PM Changeset in ghc [8575d01b]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Ian Lynagh <ian@…>
Fix the names of the libffi archives Fixes a bug which meant that they weren't getting installed.
4:22 PM Ticket #7792 (unregisterised build broken) created by igloo
Unregisterised builds are currently broken: Compiling C calls in CMM files …
2:18 PM Changeset in base [ac90f31]data-proxydbcsghc-7.8 by Max Bolingbroke <batterseapower@…>
Don't just fail if hGetBufSome is used on a non-FD: fall back on the slow path instead
10:24 AM Ticket #7791 (Bad error message when strictness annotations are used without ...) created by manzyuk
Loading the program […] into GHCi results in an error: […] The …
8:25 AM Ticket #7790 (Add dummy undefined symbols to indicate ways) created by ezyang
Currently, "way" information is not recorded into object files; it is only …
6:07 AM Commentary/Compiler/IfaceFiles edited by ezyang
(diff)

Mar 23, 2013:

11:55 PM Changeset in ghc [b30015e]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Ian Lynagh <ian@…>
Change how we handle libffi I think overall the new approach is simpler. Rather than unpacking the libffi.a and putting the .o files into our libHSrts.a, we just use the libffi.a. This change also means that when compiling programs for the dyn way, they get explicitly linked against libffi.so (rather than relying on librts.so being linked against it). This might fix a problem on FreeBSD, where programs cannot find libffi.so.
11:19 PM Ticket #7717 (Libffi errors while cross-compiling GHC 7.6.2 (Stage 1)) closed by guest
invalid
9:55 PM Changeset in ghc [f4a2796]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Ian Lynagh <ian@…>
Small configure.ac refactoring
7:45 PM Ticket #7789 (GHCI core dumps when used with VTY) created by timthelion
Load the attached script in ghci: […] (you need the vty-menu package …
1:32 PM Changeset in ghc [9548ec53]atomicsbetter-ho-cardinalitycoercibledata-kind-syntaxdecision-procedureghc-7.8ghc-lwc2ghc-parmake-gsoclate-dmdlate-lam-liftoverlapping-tyfamssimdth-newticky-for-all-letstype-natstype-nats-simplewip/Cabal-1.20wip/T8545-ghc-7.8wip/T8592wip/T8776wip/T8995-level-generalisationwip/cbv-conv-thunkwip/common-contextwip/cpr-vs-jpwip/exprAritywip/nested-cprwip/pattern-synonymswip/recurs-compatwip/simdwip/th-new by Ian Lynagh <ian@…>
Tweak the _HC_OPTS variables We no longer pass -hisuf/-osuf flags to "ghc -M". Doing so didn't really make sense with the way the -dep-suffix flags now work.
4:18 AM Ticket #7788 (Recursive type family causes <<loop>>) created by shachaf
This file: […] Dies with <<loop>>. The type family is recursive, of …
4:09 AM Ticket #7787 (modifyMVar does not restore value if callback returns error value) created by joeyadams
modifyMVar is currently implemented as follows: […] The problem is …
Note: See TracTimeline for information about the timeline view.