Phyx (Tamar Christina)
User

Projects

User does not belong to any projects.

Today

  • Clear sailing ahead.

Tomorrow

  • Clear sailing ahead.

Tuesday

  • Clear sailing ahead.

Badges

User Since
Nov 5 2014, 3:01 PM (124 w, 3 d)
Availability
Available

Recent Activity

Today

Phyx updated the Trac tickets for D3384: WIP: Implement string merging on Windows.
Sun, Mar 26, 3:28 PM
Phyx added a dependency for D3384: WIP: Implement string merging on Windows: D3383: WIP: Implement split-sections on windows..
Sun, Mar 26, 3:28 PM
Phyx added a dependent revision for D3383: WIP: Implement split-sections on windows.: D3384: WIP: Implement string merging on Windows.
Sun, Mar 26, 3:28 PM
Phyx created D3384: WIP: Implement string merging on Windows.
Sun, Mar 26, 3:25 PM
Phyx updated the Trac tickets for D3382: First update mingw-w64 packages for 8.4.
Sun, Mar 26, 3:02 PM
Phyx added a dependent revision for D3382: First update mingw-w64 packages for 8.4: D3383: WIP: Implement split-sections on windows..
Sun, Mar 26, 3:02 PM
Phyx added a dependency for D3383: WIP: Implement split-sections on windows.: D3382: First update mingw-w64 packages for 8.4.
Sun, Mar 26, 3:02 PM
Phyx updated the Trac tickets for D3383: WIP: Implement split-sections on windows..
Sun, Mar 26, 3:02 PM
Phyx added a comment to D3383: WIP: Implement split-sections on windows..

This unfortunately increases the size of the GHCi libraries.
So much that memory can't be allocated to load them. Have to
figure that out first.

Sun, Mar 26, 3:01 PM
Phyx created D3383: WIP: Implement split-sections on windows..
Sun, Mar 26, 3:01 PM
Phyx created D3382: First update mingw-w64 packages for 8.4.
Sun, Mar 26, 2:41 PM
Phyx requested changes to D3278: Enable new warning for fragile/incorrect CPP #if usage.

Looking at these again, I suspect a lot of these aren't right. If you look at includes/stg/HaskellMachRegs.h it seems like these MACHREGS_ macros are always defined. just to either 0 or 1.
By switching these from #if to #if defined(..) they're now all trivially always true.

Sun, Mar 26, 5:37 AM

Mon, Mar 20

Phyx committed rGHC713ff9207e0f: Fix Windows x86 build (authored by Phyx).
Fix Windows x86 build
Mon, Mar 20, 1:20 PM
Phyx closed D3362: Fix Windows x86 build by committing rGHC713ff9207e0f: Fix Windows x86 build.
Mon, Mar 20, 1:20 PM
Phyx updated the summary of D3362: Fix Windows x86 build.
Mon, Mar 20, 1:17 PM
Phyx updated the summary of D3362: Fix Windows x86 build.
Mon, Mar 20, 1:14 PM

Sat, Mar 18

Phyx updated the Trac tickets for D3362: Fix Windows x86 build.
Sat, Mar 18, 10:24 AM
Phyx created D3362: Fix Windows x86 build.
Sat, Mar 18, 10:23 AM

Fri, Mar 17

Phyx added a comment to D3318: GHC_STAGE1 isn't defined, use other form..

This change potentially breaks tests such as the whitespace and tarball validation tests, as now STAGE1_GHC will always point to the inplace one instead of the ones extracted to the specific folders.

Fri, Mar 17, 5:21 AM
Phyx added a comment to D3318: GHC_STAGE1 isn't defined, use other form..

Oops, forgot to press send.

Fri, Mar 17, 5:20 AM

Wed, Mar 15

Phyx added a comment to D3298: RFC: Always build GHCi libs.
In the second case, the `.o` is much faster to load than the `.a`, but we aren't building it by default.
Wed, Mar 15, 4:12 AM

Mon, Mar 13

Phyx added a comment to D3319: Fix Windows GCC driver.

I don't think I wrote this code, but my guess is that it has to do with proper handling of Ctrl-C. Can we still correctly interrupt compilation during the gcc phase without this?

Mon, Mar 13, 3:25 PM
Phyx added a comment to D2592: WIP: Dynamic linking support for Windows (Part 1/2) [skip ci].

Test

Mon, Mar 13, 1:30 PM

Sun, Mar 12

Phyx updated the diff for D3319: Fix Windows GCC driver.
  • Rebase
Sun, Mar 12, 4:34 AM
Phyx retitled D3319: Fix Windows GCC driver from Fix dwCreationFlags in GCC driver to Fix Windows GCC driver.
Sun, Mar 12, 4:16 AM

Sat, Mar 11

Phyx updated the summary of D3319: Fix Windows GCC driver.
Sat, Mar 11, 1:51 PM
Phyx updated the diff for D3319: Fix Windows GCC driver.
  • Actually fix it.
Sat, Mar 11, 1:42 PM
Phyx updated the summary of D3319: Fix Windows GCC driver.
Sat, Mar 11, 9:05 AM
Phyx planned changes to D3319: Fix Windows GCC driver.
Sat, Mar 11, 8:49 AM
Phyx updated the Trac tickets for D3319: Fix Windows GCC driver.
Sat, Mar 11, 4:52 AM
Phyx created D3319: Fix Windows GCC driver.
Sat, Mar 11, 4:49 AM

Thu, Mar 9

Phyx accepted D3304: configure.ac: Ensure that we handle case of non-present --target.

LGTM

Thu, Mar 9, 1:44 PM
Phyx abandoned D3152: Bump process version in GHC [skip ci].
Thu, Mar 9, 10:48 AM
Phyx added a comment to D3292: Update crt to latest..

Note to self. Should probably make these scripts smart enough to remove any existing older versions of libraries that it updated. So we prevent it from extracting multiple versions of the same tar to the same place.

Thu, Mar 9, 10:48 AM

Wed, Mar 8

Phyx committed rGHCa02b80fe5542: Bump haskeline submodule to fix Windows build. (authored by Phyx).
Bump haskeline submodule to fix Windows build.
Wed, Mar 8, 5:34 PM
Phyx added a comment to D3298: RFC: Always build GHCi libs.

Fair enough, the behavior is a bit counter intuitive to me though. If I understand correctly this is saying that loading the .a is faster than loading the .so?
At least from the Windows point of view this seems odd since DLLs are fully relocated and have no dangling symbols.

Wed, Mar 8, 1:31 PM
Phyx committed rGHCbd6681713a60: Win32: bump submodule to v2.5.2.0 (authored by Phyx).
Win32: bump submodule to v2.5.2.0
Wed, Mar 8, 2:19 AM
Phyx closed D3293: Win32: bump submodule to v2.5.2.0 by committing rGHCbd6681713a60: Win32: bump submodule to v2.5.2.0.
Wed, Mar 8, 2:19 AM

Tue, Mar 7

Phyx added a comment to D3292: Update crt to latest..

It fixes Trac #13384

Tue, Mar 7, 6:20 PM
Phyx created D3293: Win32: bump submodule to v2.5.2.0.
Tue, Mar 7, 4:23 PM
Phyx added a comment to D3292: Update crt to latest..

Sort of, I've updated the script so mk/get-win32-tarballs.sh download mirror will get you most things. gcc-libs doesn't seem to have it's source available on the repo and mpc is split into different sources for x86 and x86_64. The script isn't setup to handle that.

Tue, Mar 7, 4:05 PM
Phyx updated the diff for D3292: Update crt to latest..
  • update mirror script.
Tue, Mar 7, 4:05 PM
Phyx added a comment to D3292: Update crt to latest..

@bgamari run

Tue, Mar 7, 2:41 PM
Phyx updated the Trac tickets for D3292: Update crt to latest..
Tue, Mar 7, 2:40 PM
Phyx created D3292: Update crt to latest..
Tue, Mar 7, 2:39 PM

Mon, Mar 6

Phyx added a comment to D3279: Add GCC bin folder to search path..

No, I don't believe so. The path containing libwinpthread-1.dll may contain other user installed dlls,
In the case of msys2 this is /mingw64/bin, so removing it would make you not pick up dependencies you should have.
The API used gives precedence to the paths specified over those in the PATH environment. So while it can get overridden
by e.g. spawning a call to ghc from your own program which uses the same APIs, at that point I have to assume that this is the behaviour
you desired.

Mon, Mar 6, 7:00 AM
Phyx added a comment to D3279: Add GCC bin folder to search path..

No, we always want the one from the inplace gcc distro we provide.
Critically though this patch allows it to work when you don't have one in your environment.

Mon, Mar 6, 2:37 AM

Sun, Mar 5

Phyx updated the Trac tickets for D3279: Add GCC bin folder to search path..
Sun, Mar 5, 5:47 AM
Phyx created D3279: Add GCC bin folder to search path..
Sun, Mar 5, 5:46 AM
Phyx accepted D3278: Enable new warning for fragile/incorrect CPP #if usage.
Sun, Mar 5, 5:35 AM

Tue, Feb 28

Phyx committed rGHC990f182ad0cd: Fix windows build broken by D3080 (0d86aa5904e5a06c93632357122e57e4e118fd2a) (authored by Phyx).
Fix windows build broken by D3080 (0d86aa5904e5a06c93632357122e57e4e118fd2a)
Tue, Feb 28, 12:22 AM
Phyx closed D3225: Get rid of unused import warning on Windows by committing rGHC990f182ad0cd: Fix windows build broken by D3080 (0d86aa5904e5a06c93632357122e57e4e118fd2a).
Tue, Feb 28, 12:22 AM
Phyx raised a concern with rGHC0d86aa5904e5: Add support for concurrent package db access and updates.

This killed the windows build again. Harbormaster said so before the push: https://phabricator.haskell.org/harbormaster/build/21516/

Tue, Feb 28, 12:15 AM

Sun, Feb 26

Phyx added inline comments to D3082: Extend Windows runtime loader libsearch.
Sun, Feb 26, 11:29 AM

Feb 23 2017

Phyx added a comment to D3152: Bump process version in GHC [skip ci].

It would be great if we could also lift the restriction on the boot packages somehow. Maybe by using cabal to first register the new packages in the stage0 compiler in as sandbox or something?

Feb 23 2017, 11:00 AM
Phyx added a comment to D3152: Bump process version in GHC [skip ci].

Yeah, having up upgrade the dependencies first is quite a pain..

Feb 23 2017, 2:13 AM

Feb 22 2017

Phyx added a comment to D3152: Bump process version in GHC [skip ci].

ugh, that's quite unfortunate...

Feb 22 2017, 9:39 PM

Feb 21 2017

Phyx abandoned D3153: Bump win32 version in GHC.

Looks like it.

Feb 21 2017, 9:40 AM

Feb 20 2017

Phyx added a comment to D3158: Correct Windows libdir assumptions..

Ah, that looks better...

Feb 20 2017, 3:17 PM
Phyx updated the diff for D3158: Correct Windows libdir assumptions..
  • fix linux build
Feb 20 2017, 3:14 PM
Phyx added a comment to D3158: Correct Windows libdir assumptions..

It is possible, arc was being particularly difficult that day. But according to my tree

Feb 20 2017, 3:03 PM

Feb 19 2017

Phyx added a comment to D3160: Fix all broken perf tests on x64 Windows.

As can be seen https://phabricator.haskell.org/harbormaster/build/21193/

Feb 19 2017, 10:51 AM
Phyx created D3160: Fix all broken perf tests on x64 Windows.
Feb 19 2017, 10:37 AM
Phyx updated the diff for D3158: Correct Windows libdir assumptions..
  • fix linux build
Feb 19 2017, 8:38 AM
Phyx added a comment to D3158: Correct Windows libdir assumptions..

https://phabricator.haskell.org/B14010 testsuite build

Feb 19 2017, 7:25 AM
Phyx updated the Trac tickets for D3158: Correct Windows libdir assumptions..
Feb 19 2017, 7:13 AM
Phyx created D3158: Correct Windows libdir assumptions..
Feb 19 2017, 7:11 AM

Feb 18 2017

Phyx updated the Trac tickets for D3155: Load `pthreads` by default on Windows.
Feb 18 2017, 7:16 AM
Phyx added a dependency for D3155: Load `pthreads` by default on Windows: D3154: Make list of deprecated symbols on Windows weak..
Feb 18 2017, 7:15 AM
Phyx added a dependent revision for D3154: Make list of deprecated symbols on Windows weak.: D3155: Load `pthreads` by default on Windows.
Feb 18 2017, 7:15 AM
Phyx added a dependency for D3154: Make list of deprecated symbols on Windows weak.: D3028: Load dependent dlls..
Feb 18 2017, 7:15 AM
Phyx added a dependent revision for D3028: Load dependent dlls.: D3154: Make list of deprecated symbols on Windows weak..
Feb 18 2017, 7:15 AM
Phyx created D3155: Load `pthreads` by default on Windows.
Feb 18 2017, 7:14 AM
Phyx updated the Trac tickets for D3154: Make list of deprecated symbols on Windows weak..
Feb 18 2017, 7:07 AM
Phyx created D3154: Make list of deprecated symbols on Windows weak..
Feb 18 2017, 7:06 AM
Phyx added inline comments to D3136: Add -fwhole-archive-hs-libs.
Feb 18 2017, 3:38 AM
Phyx created D3153: Bump win32 version in GHC.
Feb 18 2017, 3:25 AM
Phyx retitled D3152: Bump process version in GHC [skip ci] from Bump process version to Bump process version in GHC [skip ci].
Feb 18 2017, 3:17 AM
Phyx created D3152: Bump process version in GHC [skip ci].
Feb 18 2017, 3:16 AM
Phyx updated the Trac tickets for D3030: Fix nofib benchmarks on Windows..
Feb 18 2017, 2:13 AM

Feb 16 2017

Phyx added a comment to D3141: testsuite: Move echoing commands in make invocations to VERBOSE=5.

Passing --trace already makes a test fail, passing any flag to make that's used recursively can make the tests fail. I also don't see how getting the performance output in the output unconditionally is not making the test fail.

Feb 16 2017, 2:26 AM

Feb 15 2017

Phyx planned changes to D2592: WIP: Dynamic linking support for Windows (Part 1/2) [skip ci].

Just updating with current state. after rebasing I now get some random segfaults which need to be figured out. dlltool is also way too slow so a custom import lib writer needs to be written and will leverage that to write a custom loader to solve the remaining issue of how to correctly handle const data issue introduced by GHC's dynamic linking not having been designed to work cross platform.

Feb 15 2017, 10:01 PM
Phyx updated the diff for D2592: WIP: Dynamic linking support for Windows (Part 1/2) [skip ci].
  • T5987: rebased.
  • T5987: fixing after rebase
  • T5987: Fixes after rebase.
  • T5987: reverted submodules.
  • DynLib: Fix ghc-cabal build errors and rts build error.
  • T5987: Reworking DLL building code.
  • T5987: Fix failure.
  • T5987: turn off lazy loading.
  • Dyn: remove assert code and other werr issues.
  • Disable CoreToStg asserts for DynWay for now.
  • Update submodules.
  • Shared: added gen-dll
  • Gen-Dll: Added better argument parsing
  • Calling Win32 args parser
  • GenDll: Fixed gen-dll I/O
  • Dll: Enable dynamic-too on Windows.
  • Fix some more rebase screwiness
  • Remove more dllsplit tstuff again
  • Fix execProc.
  • Disable CoreToStg warnings completely.
  • Fix -Werror
  • Correct ar script.
  • Dyn: disable import lib creation.
  • fix compile.
Feb 15 2017, 9:54 PM
Phyx updated the diff for D3028: Load dependent dlls..
  • Inlined dependency resolving into addDllHandler
  • Rebased on master
Feb 15 2017, 9:38 PM
Phyx accepted D3141: testsuite: Move echoing commands in make invocations to VERBOSE=5.

Lgtm

Feb 15 2017, 7:24 PM
Phyx added a comment to D3028: Load dependent dlls..

arc seems to currently be broken, so I can't update the patch. Unless someone knows a way to do it without arc?

Feb 15 2017, 12:38 PM

Feb 14 2017

Phyx added inline comments to D3028: Load dependent dlls..
Feb 14 2017, 9:34 AM
Phyx added inline comments to D3026: Fix ExtraSymbols jump table on Windows..
Feb 14 2017, 9:20 AM

Feb 12 2017

Phyx accepted D3129: Typecast covers entire expression to fix format warning..

Thanks @bollu !

Feb 12 2017, 5:47 AM
Phyx requested changes to D3129: Typecast covers entire expression to fix format warning..

Thanks @bollu! Just some minor comments.

Feb 12 2017, 4:52 AM

Feb 7 2017

Phyx added inline comments to D3090: Add support for concurrent package db access and updates.
Feb 7 2017, 4:51 AM

Feb 6 2017

Phyx retitled D2592: WIP: Dynamic linking support for Windows (Part 1/2) [skip ci] from WIP: Dynamic linking support for Windows (Part 1/2) [ci skip] to WIP: Dynamic linking support for Windows (Part 1/2) [skip ci].
Feb 6 2017, 11:41 AM

Feb 5 2017

Phyx updated the test plan for D3028: Load dependent dlls..
Feb 5 2017, 1:12 PM
Phyx updated the diff for D3028: Load dependent dlls..
  • make search recursive.
Feb 5 2017, 12:58 PM
Phyx added a comment to D3028: Load dependent dlls..
In D3028#90491, @Phyx wrote:

The searching for dependencies isn't recursive, which was by design.

This surprises me. I'm already not really supposed to care about symbols defined in a dependency of the library I depend on, how could I care about the difference between a direct dependency and an indirect dependency?

Feb 5 2017, 12:19 PM
Phyx updated the diff for D3028: Load dependent dlls..
  • use handle instead of filename in duplicate check
Feb 5 2017, 11:55 AM
Phyx added inline comments to D3028: Load dependent dlls..
Feb 5 2017, 11:41 AM
Phyx updated the diff for D3028: Load dependent dlls..
  • Dropping normalization.
Feb 5 2017, 11:39 AM
Phyx added a comment to D3028: Load dependent dlls..

So without the normalization we get a slightly different result because now the system dll's names won't match.

Feb 5 2017, 11:37 AM
Phyx updated the test plan for D3028: Load dependent dlls..
Feb 5 2017, 11:33 AM