Dan Gohman
9eb764b90a
Rename simd.jd to ecmascript_simd.js
...
Since that's what the upstream name of this file is. Note that we are
still using a branch of upstream rather than plain upstream.
2014-10-28 10:31:26 -07:00
Alon Zakai
1bc77f0fbf
handle some corner cases in EMULATE_FUNCTION_POINTER_CASTS, and add testing
2014-10-23 13:47:52 -07:00
Alon Zakai
eb6ac90760
EMULATE_FUNCTION_POINTER_CASTS option
2014-10-22 16:24:26 -07:00
Alon Zakai
3b656a8133
Merge branch 'incoming' into emterpreter
2014-10-17 10:55:47 -07:00
Alon Zakai
a2f12b3219
update memory growth code
2014-10-17 10:55:26 -07:00
Alon Zakai
e084763951
Merge branch 'incoming' into emterpreter
...
Conflicts:
src/preamble.js
2014-10-15 13:28:00 -07:00
Alon Zakai
170a106d59
fix swappable bug
2014-10-13 15:27:18 -07:00
Alon Zakai
2b692e14cb
SWAPPABLE_ASM_MODULE option, and finish distill
2014-10-13 14:21:35 -07:00
Alon Zakai
8cb50b7bf8
save asm module args on Module object
2014-10-13 13:18:37 -07:00
Alon Zakai
0f603c3175
remove some unneeded code for Math.min
2014-10-13 12:57:37 -07:00
Dan Gohman
0b5a03ee3d
Switch to a newer unofficial simd.js polyfill
...
Switch to the new polyfill from
https://github.com/sunfishcode/ecmascript_simd/blob/development/src/ecmascript_simd.js
for now. This anticipates several changes which are expected to be
accepted upstream.
Among other things, this contains the shuffle->swizzle rename and the
shuffleMix->shuffle rename and generalization.
I also added a workaround for the fact that SpiderMonkey isn't allowing
the polyfill to detect whether signMask is implemented, and isn't
allowing it to be overridden.
2014-10-07 10:41:49 -07:00
Dan Gohman
5657d0076e
Remove missing int32x4 comparisons for now.
...
We'll hopefully add these to the spec soon, but for now, it's not hard
to just emulate them.
2014-10-07 09:21:59 -07:00
Dan Gohman
a90e812bbf
Remove SIMD loadX, storeX, and friends for now.
...
OdinMonkey doesn't support them yet, and it'll be easy to re-introduce
them when it does.
2014-10-07 08:53:41 -07:00
Alon Zakai
0c85e6d502
option to customize GLOBAL_BASE
2014-10-03 12:26:33 -07:00
Alon Zakai
160e8049f2
option to customize GLOBAL_BASE
2014-10-03 12:18:50 -07:00
Dan Gohman
68ac0f09fe
Fix a typo.
2014-10-02 15:49:56 -07:00
Alon Zakai
e9642376f9
Merge branch 'incoming' into emterpreter
...
Conflicts:
emscripten.py
tests/test_core.py
tools/js-optimizer.js
2014-10-01 13:26:31 -07:00
Alon Zakai
21d258f329
fix static static bump calculation
2014-09-29 17:43:06 -07:00
Alon Zakai
f070262ee4
update memory growth code, and ensure replace_memory is on the top in fastcomp
2014-09-29 15:26:42 -07:00
Alon Zakai
2f57b4b2ce
fix warning on EXPORTED_FUNCTIONS, and improve test
2014-09-28 17:56:24 -07:00
Dan Gohman
722df73443
Add more SIMD API functions.
2014-09-25 14:34:25 -07:00
Dan Gohman
a0eea413fc
Asm.js validation for SIMD
2014-09-25 13:22:46 -07:00
Alon Zakai
b5ae09fd87
update heap resizing code to latest proposal
2014-09-19 11:22:04 -07:00
Alon Zakai
6d254d1f3a
simplify staticbump generation
2014-09-16 15:36:36 -07:00
Alon Zakai
8baf8346fd
experimental POINTER_MASKING option
2014-09-04 14:19:13 -07:00
Alon Zakai
a529dce792
show the fastcomp string explaining why code can't validate, when it can't; fixes #2737
2014-08-29 10:26:25 -07:00
Alon Zakai
de98a203a5
remove no longer needed rand_seed hack
2014-08-26 15:31:49 -07:00
Alon Zakai
1efbd0baa4
make stack align to 16 bytes in fastcomp, as the fastcomp backend does; #2721
2014-08-26 14:22:46 -07:00
Alon Zakai
6fb578efe9
remove old asmPrintInt and Float funcs
2014-08-22 13:09:34 -07:00
Alon Zakai
919556db48
add --closure 2 option, to run closure even on the asm.js output
2014-08-22 13:07:02 -07:00
Alon Zakai
4264dab989
add fastcomp stack overflow checks and testing
2014-08-07 16:15:54 -07:00
Bruce Mitchener
03fa2a59c1
Make no-prefixing of library functions work.
...
This makes it so that functions in JS library files can start
with '$' to signal that they shouldn't be prefixed with an '_' and
be visible to user code.
This is used in the conversion of emval and embind to JS library
format so that their internals are not exposed inappropriately.
2014-08-03 08:20:06 +07:00
Alon Zakai
8f4d5ece6d
fix bug where only ctlz or cttz is present, but not both
2014-07-29 17:47:40 -07:00
Jukka Jylänki
47c46fed9b
Fix issue #2133 with \r\r\n line endings being generated.
2014-07-29 10:01:38 +03:00
Lu Wang
394f308b6f
support ASYNCIFY
2014-07-24 17:30:25 -07:00
Alon Zakai
1fb983e3bf
do not warn on malloc not present, we do so at runtime if used (+commented code to pull in malloc and other system libs based on EXPORTS)
2014-07-21 11:39:37 -07:00
Alon Zakai
26294d2f8c
limit error output from fastcomp invocation of compiler.js, to make errors clearer, and fix other.test_warn_undefined
2014-07-11 11:28:36 -07:00
Bruce Mitchener
eb88ff745f
Let -s RELOOPER work correctly.
2014-07-08 12:43:43 +07:00
Alon Zakai
b82c5cffeb
improve assert text
2014-07-03 14:26:59 -07:00
Alon Zakai
82b259c130
fix typo
2014-07-03 13:37:26 -07:00
Alon Zakai
5aec00b0cb
guard against code execution before runtime is initialized and after it is exited
2014-07-03 13:10:39 -07:00
Alon Zakai
4d6bfe38b7
remove obsolete 'simple' mode in emscript_fast
2014-07-03 11:42:11 -07:00
Alon Zakai
d46e9600bc
set MAX_SETJMPS to fastcomp; 1.18.4
2014-05-27 11:26:40 -07:00
Alon Zakai
bf1c3dff11
wipe out emscripten.py output when it fails, so the error does not cascade weirdly; fixes #2368
2014-05-20 16:59:00 -07:00
Alon Zakai
d3e3d803ad
warn on bad exports, fixes #2338
2014-05-12 11:35:58 -07:00
Alon Zakai
2ad7b1037c
Merge pull request #2311 from fadams/fix-64bit-long-return-issue
...
Fix 64bit long return issue
2014-05-08 16:21:14 -07:00
Fraser Adams
deb49776cf
add Runtime.setTempRet0 method for symmetry
2014-05-07 20:35:40 +01:00
Alon Zakai
490a4b27d1
update setTemp* as well
2014-05-06 16:32:40 -07:00
Alon Zakai
a4e3a2c456
export SAFE_FT_MASK in non-fastcomp, to fix some asm validation errors
2014-05-05 13:25:55 -07:00
Alon Zakai
abad7b6e58
add f0 in non-fastcomp when needed
2014-05-03 16:51:26 -07:00