[pypy-commit] pypy s390x-backend: made register r13 dirty before flushing the non volatile registers to the stack,

Author: Richard Plangger Branch: s390x-backend Changeset: r81884:f5ab3026a4db Date: 2016-01-21 10:08 +0100 http://bitbucket.org/pypy/pypy/changeset/f5ab3026a4db/ Log:made register r13 dirty before flushing the non volatile registers to the stack, it seems that this cannot happen in th

[pypy-commit] pypy s390x-backend: it can happen that longevity does not contain an entry for an operation (e.g. int_mul_ovf and result is not used), then when trying to spill a variable op can be in r

Author: Richard Plangger Branch: s390x-backend Changeset: r81885:7ac200cdeecd Date: 2016-01-21 11:24 +0100 http://bitbucket.org/pypy/pypy/changeset/7ac200cdeecd/ Log:it can happen that longevity does not contain an entry for an operation (e.g. int_mul_ovf and result is not used), then

[pypy-commit] pypy s390x-backend: advancing the pointer to correctly return the value written from libffi, the reason why this does not happen on e.g. ppc bigendian 64 bit is: libffi ppc casts the ret

Author: Richard Plangger Branch: s390x-backend Changeset: r81887:08606f22af4a Date: 2016-01-21 20:37 +0100 http://bitbucket.org/pypy/pypy/changeset/08606f22af4a/ Log:advancing the pointer to correctly return the value written from libffi, the reason why this does not happen on e.g. pp

[pypy-commit] pypy s390x-backend: excluding structs from buffer advance (which would be wrong)

Author: Richard Plangger Branch: s390x-backend Changeset: r81895:60aa103a575b Date: 2016-01-21 21:04 +0100 http://bitbucket.org/pypy/pypy/changeset/60aa103a575b/ Log:excluding structs from buffer advance (which would be wrong) fixed wide_char test (tested little endian) diff --git a/

[pypy-commit] pypy s390x-backend: unicode endian test issue in pypy module

Author: Richard Plangger Branch: s390x-backend Changeset: r81900:d2af83605cf2 Date: 2016-01-22 08:36 +0100 http://bitbucket.org/pypy/pypy/changeset/d2af83605cf2/ Log:unicode endian test issue in pypy module diff --git a/pypy/module/_file/test/test_file_extra.py b/pypy/module/_file/test/test

[pypy-commit] pypy s390x-backend: removed debug print statement, return value of closure changed to long. a closure return value must fill a full ffi_arg

Author: Richard Plangger Branch: s390x-backend Changeset: r81901:4cafa238dca9 Date: 2016-01-22 08:43 +0100 http://bitbucket.org/pypy/pypy/changeset/4cafa238dca9/ Log:removed debug print statement, return value of closure changed to long. a closure return value must fill a full ffi_arg

[pypy-commit] pypy s390x-backend: big endian test issue cpyext

Author: Richard Plangger Branch: s390x-backend Changeset: r81904:43c90f2f520e Date: 2016-01-22 09:20 +0100 http://bitbucket.org/pypy/pypy/changeset/43c90f2f520e/ Log:big endian test issue cpyext diff --git a/pypy/module/cpyext/test/test_arraymodule.py b/pypy/module/cpyext/test/test_arraymod

[pypy-commit] pypy s390x-backend: big endian issue while testing cpyext (PyUnicode_DecodeUTF16)

Author: Richard Plangger Branch: s390x-backend Changeset: r81905:b1b31c094879 Date: 2016-01-22 09:34 +0100 http://bitbucket.org/pypy/pypy/changeset/b1b31c094879/ Log:big endian issue while testing cpyext (PyUnicode_DecodeUTF16) diff --git a/pypy/module/cpyext/test/test_arraymodule.py b/pypy

[pypy-commit] pypy s390x-backend: the same for the UTF32 test

Author: Richard Plangger Branch: s390x-backend Changeset: r81906:9df502324ca5 Date: 2016-01-22 09:36 +0100 http://bitbucket.org/pypy/pypy/changeset/9df502324ca5/ Log:the same for the UTF32 test diff --git a/pypy/module/cpyext/test/test_unicodeobject.py b/pypy/module/cpyext/test/test_unicode

[pypy-commit] pypy s390x-backend: endian test issue marshal, test decoded value using platform endianess (not desired for marshall module)

Author: Richard Plangger Branch: s390x-backend Changeset: r81907:359e329036d5 Date: 2016-01-22 09:59 +0100 http://bitbucket.org/pypy/pypy/changeset/359e329036d5/ Log:endian test issue marshal, test decoded value using platform endianess (not desired for marshall module) diff --git a/

[pypy-commit] pypy s390x-backend: numpy dtype fixes in the test suite

Author: Richard Plangger Branch: s390x-backend Changeset: r81908:6b28800745be Date: 2016-01-22 10:21 +0100 http://bitbucket.org/pypy/pypy/changeset/6b28800745be/ Log:numpy dtype fixes in the test suite diff --git a/pypy/module/marshal/test/test_marshalimpl.py b/pypy/module/marshal/test/test

[pypy-commit] pypy s390x-backend: two more dtype tests fixed (endian issues)

Author: Richard Plangger Branch: s390x-backend Changeset: r81909:272b467ba7dd Date: 2016-01-22 10:24 +0100 http://bitbucket.org/pypy/pypy/changeset/272b467ba7dd/ Log:two more dtype tests fixed (endian issues) diff --git a/pypy/module/micronumpy/test/test_dtypes.py b/pypy/module/micronumpy/t

[pypy-commit] pypy s390x-backend: half way through the ndarray tests (endian issues)

Author: Richard Plangger Branch: s390x-backend Changeset: r81910:d1a60e575946 Date: 2016-01-22 10:56 +0100 http://bitbucket.org/pypy/pypy/changeset/d1a60e575946/ Log:half way through the ndarray tests (endian issues) diff --git a/pypy/module/micronumpy/test/test_ndarray.py b/pypy/module/mic

[pypy-commit] pypy s390x-backend: and the other part of the bigendian issues (micronumpy tests)

Author: Richard Plangger Branch: s390x-backend Changeset: r81911:47a85e21bb1b Date: 2016-01-22 12:00 +0100 http://bitbucket.org/pypy/pypy/changeset/47a85e21bb1b/ Log:and the other part of the bigendian issues (micronumpy tests) diff --git a/pypy/module/micronumpy/test/test_ndarray.py b/pypy

[pypy-commit] pypy s390x-backend: fixed callsite of clibffi with the same big endian issues as found yesterday evening

Author: Richard Plangger Branch: s390x-backend Changeset: r81912:6840459f9b22 Date: 2016-01-22 12:51 +0100 http://bitbucket.org/pypy/pypy/changeset/6840459f9b22/ Log:fixed callsite of clibffi with the same big endian issues as found yesterday evening diff --git a/pypy/module/micronum

[pypy-commit] pypy s390x-backend: macros (e.g. WCOREDUMP) got parameter type Signed, on little endian this does not make a difference, but it does on big endian. changed to rffi.INT

Author: Richard Plangger Branch: s390x-backend Changeset: r81914:43db866bfbb3 Date: 2016-01-22 14:56 +0100 http://bitbucket.org/pypy/pypy/changeset/43db866bfbb3/ Log:macros (e.g. WCOREDUMP) got parameter type Signed, on little endian this does not make a difference, but it does on big

[pypy-commit] pypy s390x-backend: added comment to my last commit

Author: Richard Plangger Branch: s390x-backend Changeset: r81922:dcf9b353608e Date: 2016-01-22 18:03 +0100 http://bitbucket.org/pypy/pypy/changeset/dcf9b353608e/ Log:added comment to my last commit diff --git a/rpython/rlib/rposix.py b/rpython/rlib/rposix.py --- a/rpython/rlib/rposix.py +++

[pypy-commit] pypy s390x-backend: removed lldebug properties from test translation (left them for debugging), fixed a bug when Const arg is provided to force_allocate_reg, it could have consumed on of

Author: Richard Plangger Branch: s390x-backend Changeset: r81927:3430d1f200fa Date: 2016-01-25 11:02 +0100 http://bitbucket.org/pypy/pypy/changeset/3430d1f200fa/ Log:removed lldebug properties from test translation (left them for debugging), fixed a bug when Const arg is provided to

[pypy-commit] pypy s390x-backend: rewritten many calls to use one stack frame less

Author: Richard Plangger Branch: s390x-backend Changeset: r81928:6a1b2984c003 Date: 2016-01-25 13:31 +0100 http://bitbucket.org/pypy/pypy/changeset/6a1b2984c003/ Log:rewritten many calls to use one stack frame less diff --git a/rpython/jit/backend/zarch/arch.py b/rpython/jit/backend/zarch/a

[pypy-commit] pypy s390x-backend: fixed problem in build_propagate_exception_path, value was loaded into wrong register (that was overwritten after that immediatley)

Author: Richard Plangger Branch: s390x-backend Changeset: r81929:2c1aa1edc33a Date: 2016-01-25 13:54 +0100 http://bitbucket.org/pypy/pypy/changeset/2c1aa1edc33a/ Log:fixed problem in build_propagate_exception_path, value was loaded into wrong register (that was overwritten after that

[pypy-commit] pypy s390x-backend: added missing parameter to LMG call (refactoring issues :)

Author: Richard Plangger Branch: s390x-backend Changeset: r81932:0e52bea859f2 Date: 2016-01-25 14:04 +0100 http://bitbucket.org/pypy/pypy/changeset/0e52bea859f2/ Log:added missing parameter to LMG call (refactoring issues :) diff --git a/rpython/jit/backend/zarch/codebuilder.py b/rpython/ji

[pypy-commit] pypy s390x-backend: due to the refacotring the variable saving went to the wrong stack frame in wb_slowpath (now takes the right one)

Author: Richard Plangger Branch: s390x-backend Changeset: r81933:adce895c17e8 Date: 2016-01-25 15:54 +0100 http://bitbucket.org/pypy/pypy/changeset/adce895c17e8/ Log:due to the refacotring the variable saving went to the wrong stack frame in wb_slowpath (now takes the right one) diff

[pypy-commit] pypy s390x-backend: removed longlong from s390x from cpu

Author: Richard Plangger Branch: s390x-backend Changeset: r81943:47d8fe0b9435 Date: 2016-01-26 08:42 +0100 http://bitbucket.org/pypy/pypy/changeset/47d8fe0b9435/ Log:removed longlong from s390x from cpu diff --git a/rpython/jit/backend/detect_cpu.py b/rpython/jit/backend/detect_cpu.py --- a

[pypy-commit] pypy s390x-backend: merged default

Author: Richard Plangger Branch: s390x-backend Changeset: r81944:c6029ee6abcc Date: 2016-01-26 08:47 +0100 http://bitbucket.org/pypy/pypy/changeset/c6029ee6abcc/ Log:merged default diff --git a/pypy/doc/whatsnew-head.rst b/pypy/doc/whatsnew-head.rst --- a/pypy/doc/whatsnew-head.rst +++ b/pyp

[pypy-commit] pypy s390x-backend: removed print statements, typo, removed push floats from stacklet because there is no need to save them

Author: Richard Plangger Branch: s390x-backend Changeset: r81946:62aeb645010b Date: 2016-01-26 09:07 +0100 http://bitbucket.org/pypy/pypy/changeset/62aeb645010b/ Log:removed print statements, typo, removed push floats from stacklet because there is no need to save them diff --git a/r

[pypy-commit] pypy s390x-backend: simplified guard_class

Author: Richard Plangger Branch: s390x-backend Changeset: r81947:1589c0b36069 Date: 2016-01-26 09:36 +0100 http://bitbucket.org/pypy/pypy/changeset/1589c0b36069/ Log:simplified guard_class diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpython/jit/backend/zarch/opassembler.py --- a

[pypy-commit] pypy s390x-backend: gotcha, guard nonnull class was implemented incorrectly (substracted value that should have been shifted 16 bits to the left)

Author: Richard Plangger Branch: s390x-backend Changeset: r81948:acf6ac9d0d3d Date: 2016-01-26 14:01 +0100 http://bitbucket.org/pypy/pypy/changeset/acf6ac9d0d3d/ Log:gotcha, guard nonnull class was implemented incorrectly (substracted value that should have been shifted 16 bits to the

[pypy-commit] pypy s390x-backend: merged head

Author: Richard Plangger Branch: s390x-backend Changeset: r81950:3d78eb62b0e5 Date: 2016-01-26 16:44 +0100 http://bitbucket.org/pypy/pypy/changeset/3d78eb62b0e5/ Log:merged head diff --git a/rpython/jit/backend/zarch/assembler.py b/rpython/jit/backend/zarch/assembler.py --- a/rpython/jit/ba

[pypy-commit] pypy s390x-backend: fixed translation issue. reverted the stacklet changes (did not pass test)

Author: Richard Plangger Branch: s390x-backend Changeset: r81949:fb7bbc354de2 Date: 2016-01-26 16:40 +0100 http://bitbucket.org/pypy/pypy/changeset/fb7bbc354de2/ Log:fixed translation issue. reverted the stacklet changes (did not pass test) diff --git a/rpython/jit/backend/zarch/rega

[pypy-commit] pypy s390x-backend: added ztranslation call assembler test and some more (hurray, one of them fails and this might be the bug I'm searching for)

Author: Richard Plangger Branch: s390x-backend Changeset: r81983:228ce776f7f6 Date: 2016-01-27 20:46 +0100 http://bitbucket.org/pypy/pypy/changeset/228ce776f7f6/ Log:added ztranslation call assembler test and some more (hurray, one of them fails and this might be the bug I'm searching

[pypy-commit] pypy s390x-backend: removed size info from call builder when assembling call_assembler (it is not needed),

Author: Richard Plangger Branch: s390x-backend Changeset: r81992:e6d4a987b802 Date: 2016-01-28 11:32 +0100 http://bitbucket.org/pypy/pypy/changeset/e6d4a987b802/ Log:removed size info from call builder when assembling call_assembler (it is not needed), removed several not necessary st

[pypy-commit] pypy s390x-backend: added SGRK instruction, now the stack overflow is correctly detected and the exception is thrown!

Author: Richard Plangger Branch: s390x-backend Changeset: r81993:394d3ce64a00 Date: 2016-01-28 14:12 +0100 http://bitbucket.org/pypy/pypy/changeset/394d3ce64a00/ Log:added SGRK instruction, now the stack overflow is correctly detected and the exception is thrown! diff --git a/rpython

[pypy-commit] pypy s390x-backend: extended the stack limit of s390x

Author: Richard Plangger Branch: s390x-backend Changeset: r81994:eaa7b84cc9cd Date: 2016-01-28 14:21 +0100 http://bitbucket.org/pypy/pypy/changeset/eaa7b84cc9cd/ Log:extended the stack limit of s390x diff --git a/rpython/translator/c/src/stack.h b/rpython/translator/c/src/stack.h --- a/rpyth

[pypy-commit] pypy s390x-backend: greater equal is not equal, this basically invoked realloc frame EVERY time the assembler location was visited

Author: Richard Plangger Branch: s390x-backend Changeset: r81998:2f0e451a18e2 Date: 2016-01-28 21:35 +0100 http://bitbucket.org/pypy/pypy/changeset/2f0e451a18e2/ Log:greater equal is not equal, this basically invoked realloc frame EVERY time the assembler location was visited diff --

[pypy-commit] pypy s390x-backend: adapted test to match the expected assembly after the changes

Author: Richard Plangger Branch: s390x-backend Changeset: r81999:e13400c9ff8a Date: 2016-01-28 22:08 +0100 http://bitbucket.org/pypy/pypy/changeset/e13400c9ff8a/ Log:adapted test to match the expected assembly after the changes diff --git a/rpython/jit/backend/zarch/test/test_runner.py b/rp

[pypy-commit] pypy s390x-backend: splitting tests in zarch backend, should speed up things quite a bit

Author: Richard Plangger Branch: s390x-backend Changeset: r82002:986957b12f49 Date: 2016-01-29 10:30 +0100 http://bitbucket.org/pypy/pypy/changeset/986957b12f49/ Log:splitting tests in zarch backend, should speed up things quite a bit diff --git a/pypy/testrunner_cfg.py b/pypy/testrunner_cfg

[pypy-commit] pypy s390x-backend: catchup with default

Author: Richard Plangger Branch: s390x-backend Changeset: r82023:f3b48b1063f6 Date: 2016-02-01 09:11 +0100 http://bitbucket.org/pypy/pypy/changeset/f3b48b1063f6/ Log:catchup with default diff too long, truncating to 2000 out of 2165 lines diff --git a/pypy/doc/faq.rst b/pypy/doc/faq.rst ---

[pypy-commit] pypy s390x-backend: replaced test case parsetuple(..."i"...) with "l",

Author: Richard Plangger Branch: s390x-backend Changeset: r82030:7be63f4c3bc8 Date: 2016-02-01 13:08 +0100 http://bitbucket.org/pypy/pypy/changeset/7be63f4c3bc8/ Log:replaced test case parsetuple(..."i"...) with "l", memory fence was place one instruction too late, replaced checkpoint

[pypy-commit] pypy s390x-backend: shadowold was a pointer not the value (which it should have been) and the comparison compared the pointers not the values

Author: Richard Plangger Branch: s390x-backend Changeset: r82031:5fe38e24272c Date: 2016-02-01 13:52 +0100 http://bitbucket.org/pypy/pypy/changeset/5fe38e24272c/ Log:shadowold was a pointer not the value (which it should have been) and the comparison compared the pointers not the valu

[pypy-commit] pypy s390x-backend: ups, need to call method load on code builder, not assembler

Author: Richard Plangger Branch: s390x-backend Changeset: r82032:035eb5778ead Date: 2016-02-01 13:56 +0100 http://bitbucket.org/pypy/pypy/changeset/035eb5778ead/ Log:ups, need to call method load on code builder, not assembler diff --git a/rpython/jit/backend/zarch/callbuilder.py b/rpython/

[pypy-commit] pypy s390x-backend: format and exchanged some registers, no breaking changes

Author: Richard Plangger Branch: s390x-backend Changeset: r82033:46f5a9e87aae Date: 2016-02-01 14:21 +0100 http://bitbucket.org/pypy/pypy/changeset/46f5a9e87aae/ Log:format and exchanged some registers, no breaking changes diff --git a/rpython/jit/backend/zarch/callbuilder.py b/rpython/jit/

[pypy-commit] pypy s390x-backend: for test case zrpy_gc_direct seems that it can prove that float storage is always 0, thus it will not compile for those two

Author: Richard Plangger Branch: s390x-backend Changeset: r82036:72e40b33ebc5 Date: 2016-02-01 17:43 +0100 http://bitbucket.org/pypy/pypy/changeset/72e40b33ebc5/ Log:for test case zrpy_gc_direct seems that it can prove that float storage is always 0, thus it will not compile for those

[pypy-commit] pypy s390x-backend: size_alignment_pos now supports bitfields for big endian platforms, some minor refactorings and simplifications

Author: Richard Plangger Branch: s390x-backend Changeset: r82041:254804e6eb46 Date: 2016-02-02 13:38 +0100 http://bitbucket.org/pypy/pypy/changeset/254804e6eb46/ Log:size_alignment_pos now supports bitfields for big endian platforms, some minor refactorings and simplifications diff -

[pypy-commit] pypy s390x-backend: reading level=2 cache for estimation size of nursery

Author: Richard Plangger Branch: s390x-backend Changeset: r82042:037ac225f6c1 Date: 2016-02-02 17:17 +0100 http://bitbucket.org/pypy/pypy/changeset/037ac225f6c1/ Log:reading level=2 cache for estimation size of nursery diff --git a/rpython/jit/backend/ppc/callbuilder.py b/rpython/jit/backen

[pypy-commit] pypy s390x-backend: added an explicit memory barrier after compare and swap

Author: Richard Plangger Branch: s390x-backend Changeset: r82043:897d3634e91c Date: 2016-02-02 18:42 +0100 http://bitbucket.org/pypy/pypy/changeset/897d3634e91c/ Log:added an explicit memory barrier after compare and swap diff --git a/rpython/jit/backend/zarch/callbuilder.py b/rpython/jit/b

[pypy-commit] pypy s390x-backend: replaced tab with 8 spaces

Author: Richard Plangger Branch: s390x-backend Changeset: r82050:067d1f873658 Date: 2016-02-03 08:24 +0100 http://bitbucket.org/pypy/pypy/changeset/067d1f873658/ Log:replaced tab with 8 spaces diff --git a/rpython/memory/gc/test/test_env.py b/rpython/memory/gc/test/test_env.py --- a/rpython

[pypy-commit] pypy s390x-backend: removed checkpoint/serialization points, zarch is sequentially consistent. some small test fixes

Author: Richard Plangger Branch: s390x-backend Changeset: r82051:01a7007c80c5 Date: 2016-02-03 08:35 +0100 http://bitbucket.org/pypy/pypy/changeset/01a7007c80c5/ Log:removed checkpoint/serialization points, zarch is sequentially consistent. some small test fixes diff --git a/pypy/mod

[pypy-commit] pypy s390x-backend: disabled blocking tests

Author: Richard Plangger Branch: s390x-backend Changeset: r82053:074033f70453 Date: 2016-02-03 11:01 +0100 http://bitbucket.org/pypy/pypy/changeset/074033f70453/ Log:disabled blocking tests diff --git a/pypy/module/__pypy__/test/test_signal.py b/pypy/module/__pypy__/test/test_signal.py ---

[pypy-commit] pypy s390x-backend: modified test to see if an empty print differs from a print with newline

Author: Richard Plangger Branch: s390x-backend Changeset: r82052:18bd2d236f85 Date: 2016-02-03 10:52 +0100 http://bitbucket.org/pypy/pypy/changeset/18bd2d236f85/ Log:modified test to see if an empty print differs from a print with newline diff --git a/pypy/module/select/test/test_sel

[pypy-commit] pypy s390x-backend: added zEC12 to assembler command for test

Author: Richard Plangger Branch: s390x-backend Changeset: r82055:a32a0a3922a4 Date: 2016-02-03 15:00 +0100 http://bitbucket.org/pypy/pypy/changeset/a32a0a3922a4/ Log:added zEC12 to assembler command for test diff --git a/rpython/jit/backend/zarch/assembler.py b/rpython/jit/backend/zarch/ass

[pypy-commit] pypy s390x-backend: do not execute auto assembler on 5 args + (takes far too long)

Author: Richard Plangger Branch: s390x-backend Changeset: r82056:e9fe4f2225ca Date: 2016-02-03 15:08 +0100 http://bitbucket.org/pypy/pypy/changeset/e9fe4f2225ca/ Log:do not execute auto assembler on 5 args + (takes far too long) diff --git a/rpython/jit/backend/zarch/test/test_auto_encoding.

[pypy-commit] pypy s390x-backend: big endian issue in cppyy test case

Author: Richard Plangger Branch: s390x-backend Changeset: r82057:f49815ceed3a Date: 2016-02-03 15:19 +0100 http://bitbucket.org/pypy/pypy/changeset/f49815ceed3a/ Log:big endian issue in cppyy test case diff --git a/pypy/module/cppyy/src/dummy_backend.cxx b/pypy/module/cppyy/src/dummy_backen

[pypy-commit] pypy s390x-backend: disabled more in test signal (bbot seems to still timeout...), simplifications in the malloc assembly operations

Author: Richard Plangger Branch: s390x-backend Changeset: r82054:b2ed3adc63e7 Date: 2016-02-03 13:55 +0100 http://bitbucket.org/pypy/pypy/changeset/b2ed3adc63e7/ Log:disabled more in test signal (bbot seems to still timeout...), simplifications in the malloc assembly operations diff

[pypy-commit] pypy s390x-backend: fixed translation issue.

Author: Richard Plangger Branch: s390x-backend Changeset: r82058:9fabf4399c18 Date: 2016-02-03 16:13 +0100 http://bitbucket.org/pypy/pypy/changeset/9fabf4399c18/ Log:fixed translation issue. diff --git a/rpython/jit/backend/zarch/assembler.py b/rpython/jit/backend/zarch/assembler.py --- a/r

[pypy-commit] pypy s390x-backend: critical bugfix in gc writebarrier fastpath, simplifications and a test

Author: Richard Plangger Branch: s390x-backend Changeset: r82059:822ac80ddd23 Date: 2016-02-03 17:43 +0100 http://bitbucket.org/pypy/pypy/changeset/822ac80ddd23/ Log:critical bugfix in gc writebarrier fastpath, simplifications and a test diff --git a/rpython/jit/backend/zarch/assembl

[pypy-commit] pypy s390x-backend: give the saved registers r10, r11, r12, r2, f0 some space on the stack. this prevents overwriting of values if they are set in the calling function

Author: Richard Plangger Branch: s390x-backend Changeset: r82062:38a946dc43e3 Date: 2016-02-03 18:49 +0100 http://bitbucket.org/pypy/pypy/changeset/38a946dc43e3/ Log:give the saved registers r10,r11,r12,r2,f0 some space on the stack. this prevents overwriting of values if they are set

[pypy-commit] pypy s390x-backend: cond_call_gc_wb_array can now not trash a volatile register. wrong allocation now takes a non volatile register!

Author: Richard Plangger Branch: s390x-backend Changeset: r82063:621a42ebea23 Date: 2016-02-03 20:19 +0100 http://bitbucket.org/pypy/pypy/changeset/621a42ebea23/ Log:cond_call_gc_wb_array can now not trash a volatile register. wrong allocation now takes a non volatile register! diff

[pypy-commit] pypy s390x-backend: minor fix. loading from pool does not work

Author: Richard Plangger Branch: s390x-backend Changeset: r82064:0ef7606ff55f Date: 2016-02-03 22:21 +0100 http://bitbucket.org/pypy/pypy/changeset/0ef7606ff55f/ Log:minor fix. loading from pool does not work diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpython/jit/backend/zarch/

[pypy-commit] pypy s390x-backend: STC can only tak 12 bits imm, this is wrong (use STCY again), fixed pool issue

Author: Richard Plangger Branch: s390x-backend Changeset: r82065:785c6853c253 Date: 2016-02-03 22:44 +0100 http://bitbucket.org/pypy/pypy/changeset/785c6853c253/ Log:STC can only tak 12 bits imm, this is wrong (use STCY again), fixed pool issue diff --git a/rpython/jit/backend/zarch/

[pypy-commit] pypy s390x-backend: added two more tests to be sure the assembler behaves correctly

Author: Richard Plangger Branch: s390x-backend Changeset: r82068:ba08e7241de3 Date: 2016-02-04 09:25 +0100 http://bitbucket.org/pypy/pypy/changeset/ba08e7241de3/ Log:added two more tests to be sure the assembler behaves correctly diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpyth

[pypy-commit] pypy s390x-backend: proper fix for the volatile reg. provided to gc write barrier array

Author: Richard Plangger Branch: s390x-backend Changeset: r82067:71aa29c678e8 Date: 2016-02-04 08:44 +0100 http://bitbucket.org/pypy/pypy/changeset/71aa29c678e8/ Log:proper fix for the volatile reg. provided to gc write barrier array diff --git a/rpython/jit/backend/zarch/opassembler.py b/r

[pypy-commit] pypy s390x-backend: an edge case in call release gil that could overwrite values for registers r8-r13, because the stack is not decremented accordingly

Author: Richard Plangger Branch: s390x-backend Changeset: r82069:f5ccdea0 Date: 2016-02-04 10:04 +0100 http://bitbucket.org/pypy/pypy/changeset/f5ccdea0/ Log:an edge case in call release gil that could overwrite values for registers r8-r13, because the stack is not decremented

[pypy-commit] pypy s390x-backend: place the saved register (loc_index) at a valid stack position!

Author: Richard Plangger Branch: s390x-backend Changeset: r82070:8bc88d808cbf Date: 2016-02-04 10:19 +0100 http://bitbucket.org/pypy/pypy/changeset/8bc88d808cbf/ Log:place the saved register (loc_index) at a valid stack position! diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpyth

[pypy-commit] pypy s390x-backend: did not consider normal gc write barrier, crashes with index out of bounds!

Author: Richard Plangger Branch: s390x-backend Changeset: r82071:7694772ae0c5 Date: 2016-02-04 10:26 +0100 http://bitbucket.org/pypy/pypy/changeset/7694772ae0c5/ Log:did not consider normal gc write barrier, crashes with index out of bounds! diff --git a/rpython/jit/backend/zarch/opa

[pypy-commit] pypy s390x-backend: translation issue

Author: Richard Plangger Branch: s390x-backend Changeset: r82072:8475d0a75240 Date: 2016-02-04 11:12 +0100 http://bitbucket.org/pypy/pypy/changeset/8475d0a75240/ Log:translation issue diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpython/jit/backend/zarch/opassembler.py --- a/rpyt

[pypy-commit] pypy s390x-backend: allocating the back chain correctly, one problem when hitting a memory error, stack size would have been incorrect!

Author: Richard Plangger Branch: s390x-backend Changeset: r82085:719886e6865e Date: 2016-02-05 13:30 +0100 http://bitbucket.org/pypy/pypy/changeset/719886e6865e/ Log:allocating the back chain correctly, one problem when hitting a memory error, stack size would have been incorrect! di

[pypy-commit] pypy s390x-backend: backchain was not correct (now it is, tested it), flush_cc must write 1, it must not be something != 1

Author: Richard Plangger Branch: s390x-backend Changeset: r82086:e108cd4706e3 Date: 2016-02-05 14:59 +0100 http://bitbucket.org/pypy/pypy/changeset/e108cd4706e3/ Log:backchain was not correct (now it is, tested it), flush_cc must write 1, it must not be something != 1 diff --git a/rp

[pypy-commit] pypy s390x-backend: must branch if not equal (error introduced in the last 2 commits)

Author: Richard Plangger Branch: s390x-backend Changeset: r82087:b95937cb9eb3 Date: 2016-02-05 15:04 +0100 http://bitbucket.org/pypy/pypy/changeset/b95937cb9eb3/ Log:must branch if not equal (error introduced in the last 2 commits) diff --git a/rpython/jit/backend/zarch/assembler.py b/rpyth

[pypy-commit] pypy s390x-backend: test checks that trace is formed for a little endian read, forcing to read little endian on s390x

Author: Richard Plangger Branch: s390x-backend Changeset: r82088:cfb1f88ae10c Date: 2016-02-05 16:58 +0100 http://bitbucket.org/pypy/pypy/changeset/cfb1f88ae10c/ Log:test checks that trace is formed for a little endian read, forcing to read little endian on s390x diff --git a/pypy/mo

[pypy-commit] pypy s390x-backend: fixed a bug where pool did not allocate space for constant parameter (gc_load_indexed, and store)

Author: Richard Plangger Branch: s390x-backend Changeset: r82090:ed27d3e4b9d8 Date: 2016-02-05 17:23 +0100 http://bitbucket.org/pypy/pypy/changeset/ed27d3e4b9d8/ Log:fixed a bug where pool did not allocate space for constant parameter (gc_load_indexed, and store) diff --git a/rpython

[pypy-commit] pypy s390x-backend: call int should provide rffi.INT instead of rffi.SIGNED, this works on little endian, but not big

Author: Richard Plangger Branch: s390x-backend Changeset: r82094:ca18f251c6bf Date: 2016-02-05 20:46 +0100 http://bitbucket.org/pypy/pypy/changeset/ca18f251c6bf/ Log:call int should provide rffi.INT instead of rffi.SIGNED, this works on little endian, but not big diff --git a/rpython

[pypy-commit] pypy s390x-backend: ffi call fixed in deprecated api that is still used (fix before that was not sufficient), fixed legacy tests test_libffi

Author: Richard Plangger Branch: s390x-backend Changeset: r82095:d40d932f8349 Date: 2016-02-05 22:37 +0100 http://bitbucket.org/pypy/pypy/changeset/d40d932f8349/ Log:ffi call fixed in deprecated api that is still used (fix before that was not sufficient), fixed legacy tests test_libff

[pypy-commit] pypy s390x-backend: refactoring issue

Author: Richard Plangger Branch: s390x-backend Changeset: r82096:a84f5c5bcec6 Date: 2016-02-05 22:42 +0100 http://bitbucket.org/pypy/pypy/changeset/a84f5c5bcec6/ Log:refactoring issue diff --git a/rpython/rlib/libffi.py b/rpython/rlib/libffi.py --- a/rpython/rlib/libffi.py +++ b/rpython/rlib

[pypy-commit] pypy s390x-backend: adding signed char to the list of bigendian conversions in rawffi.alt module

Author: Richard Plangger Branch: s390x-backend Changeset: r82113:398668348638 Date: 2016-02-08 08:25 +0100 http://bitbucket.org/pypy/pypy/changeset/398668348638/ Log:adding signed char to the list of bigendian conversions in rawffi.alt module diff --git a/rpython/rlib/libffi.py b/rpy

[pypy-commit] pypy s390x-backend: LGR should have been AGR

Author: Richard Plangger Branch: s390x-backend Changeset: r82114:e6ff00efc05f Date: 2016-02-08 08:43 +0100 http://bitbucket.org/pypy/pypy/changeset/e6ff00efc05f/ Log:LGR should have been AGR diff --git a/rpython/jit/backend/zarch/opassembler.py b/rpython/jit/backend/zarch/opassembler.py ---

[pypy-commit] pypy s390x-backend: LE set OF flag, GE set OF flag, replace guard jump from a 12bit jump to 20bit jump (relative)

Author: Richard Plangger Branch: s390x-backend Changeset: r82123:eda46a48b3de Date: 2016-02-09 11:52 +0100 http://bitbucket.org/pypy/pypy/changeset/eda46a48b3de/ Log:LE set OF flag, GE set OF flag, replace guard jump from a 12bit jump to 20bit jump (relative) diff --git a/rpython/jit

[pypy-commit] pypy s390x-backend: reviewed guarding, made negate array lookup instead of fiddling with the mask bit

Author: Richard Plangger Branch: s390x-backend Changeset: r82127:ce150b6b822b Date: 2016-02-09 18:16 +0100 http://bitbucket.org/pypy/pypy/changeset/ce150b6b822b/ Log:reviewed guarding, made negate array lookup instead of fiddling with the mask bit diff --git a/rpython/jit/backend/zar

[pypy-commit] pypy s390x-backend: fixed test_runner asmlen, since GE has the overflow bit set this test fails

Author: Richard Plangger Branch: s390x-backend Changeset: r82128:42e7257e5777 Date: 2016-02-09 18:47 +0100 http://bitbucket.org/pypy/pypy/changeset/42e7257e5777/ Log:fixed test_runner asmlen, since GE has the overflow bit set this test fails diff --git a/rpython/jit/backend/zarch/tes

[pypy-commit] pypy s390x-backend: catchup with default

Author: Richard Plangger Branch: s390x-backend Changeset: r82137:506ffad57f5e Date: 2016-02-10 09:01 +0100 http://bitbucket.org/pypy/pypy/changeset/506ffad57f5e/ Log:catchup with default diff too long, truncating to 2000 out of 4971 lines diff --git a/.hgignore b/.hgignore --- a/.hgignore +

[pypy-commit] pypy s390x-backend: ups, f1, f3, f5, f7 are volatiles, but where not added to the list of volatiles

Author: Richard Plangger Branch: s390x-backend Changeset: r82139:67c6fa68e2b9 Date: 2016-02-10 11:18 +0100 http://bitbucket.org/pypy/pypy/changeset/67c6fa68e2b9/ Log:ups, f1,f3,f5,f7 are volatiles, but where not added to the list of volatiles diff --git a/rpython/jit/backend/zarch/re

[pypy-commit] pypy s390x-backend: saving f8 through f15 before entering the jit and restoring it before exiting it. (ABI demands this)

Author: Richard Plangger Branch: s390x-backend Changeset: r82140:6cf6a1b5353a Date: 2016-02-10 12:43 +0100 http://bitbucket.org/pypy/pypy/changeset/6cf6a1b5353a/ Log:saving f8 through f15 before entering the jit and restoring it before exiting it. (ABI demands this) diff --git a/rpyt

[pypy-commit] pypy s390x-backend: putting thread local to the right position (after moving the whole frame down)

Author: Richard Plangger Branch: s390x-backend Changeset: r82141:5bae6a642fe5 Date: 2016-02-10 13:02 +0100 http://bitbucket.org/pypy/pypy/changeset/5bae6a642fe5/ Log:putting thread local to the right position (after moving the whole frame down) diff --git a/rpython/jit/backend/zarch/

[pypy-commit] buildbot default: s390x app level and jit suite use python for building

Author: Richard Plangger Branch: Changeset: r985:d4cc08767a0e Date: 2016-02-10 13:58 +0100 http://bitbucket.org/pypy/buildbot/changeset/d4cc08767a0e/ Log:s390x app level and jit suite use python for building diff --git a/bot2/pypybuildbot/master.py b/bot2/pypybuildbot/master.py --- a/bot2/p

[pypy-commit] buildbot default: merged default

Author: Richard Plangger Branch: Changeset: r986:928b30c7fc85 Date: 2016-02-10 13:59 +0100 http://bitbucket.org/pypy/buildbot/changeset/928b30c7fc85/ Log:merged default diff --git a/bot2/pypybuildbot/builds.py b/bot2/pypybuildbot/builds.py --- a/bot2/pypybuildbot/builds.py +++ b/bot2/pypybu

[pypy-commit] pypy s390x-backend: _rawffi bigendian issue in callbacks, callback writes narrow integer directly to MSB of 64 bit value on s390x (wrong when value is passed along)

Author: Richard Plangger Branch: s390x-backend Changeset: r82149:74ebd8669f96 Date: 2016-02-10 18:07 +0100 http://bitbucket.org/pypy/pypy/changeset/74ebd8669f96/ Log:_rawffi bigendian issue in callbacks, callback writes narrow integer directly to MSB of 64 bit value on s390x (wrong wh

[pypy-commit] pypy s390x-backend: translation issue

Author: Richard Plangger Branch: s390x-backend Changeset: r82148:8ca0c94c91ac Date: 2016-02-10 13:21 +0100 http://bitbucket.org/pypy/pypy/changeset/8ca0c94c91ac/ Log:translation issue diff --git a/rpython/jit/backend/zarch/assembler.py b/rpython/jit/backend/zarch/assembler.py --- a/rpython/

[pypy-commit] pypy s390x-backend: added assert, fixed test (big endian issue with unions)

Author: Richard Plangger Branch: s390x-backend Changeset: r82155:9638a3d105fa Date: 2016-02-11 09:15 +0100 http://bitbucket.org/pypy/pypy/changeset/9638a3d105fa/ Log:added assert, fixed test (big endian issue with unions) diff --git a/pypy/module/test_lib_pypy/ctypes_tests/test_unions.py b/

[pypy-commit] pypy s390x-backend: fixed two tests where the endian could not match the sequence of operations

Author: Richard Plangger Branch: s390x-backend Changeset: r82156:5040efcfecd0 Date: 2016-02-11 09:31 +0100 http://bitbucket.org/pypy/pypy/changeset/5040efcfecd0/ Log:fixed two tests where the endian could not match the sequence of operations diff --git a/pypy/module/pypyjit/test_pypy

[pypy-commit] pypy s390x-backend: misaligned is fine for s390x (rawstorage)

Author: Richard Plangger Branch: s390x-backend Changeset: r82157:c3ae24173bca Date: 2016-02-11 10:20 +0100 http://bitbucket.org/pypy/pypy/changeset/c3ae24173bca/ Log:misaligned is fine for s390x (rawstorage) diff --git a/rpython/rlib/rawstorage.py b/rpython/rlib/rawstorage.py --- a/rpython/r

[pypy-commit] pypy s390x-backend: using gc_load_r for reading a constptr since index scale and offset are constant and added to the index in rewrite, ndarray test that would fail on little endian (did

Author: Richard Plangger Branch: s390x-backend Changeset: r82166:f535e775dd54 Date: 2016-02-11 16:54 +0100 http://bitbucket.org/pypy/pypy/changeset/f535e775dd54/ Log:using gc_load_r for reading a constptr since index scale and offset are constant and added to the index in rewrite, nda

[pypy-commit] pypy s390x-backend: fixed test. emitting gc_load_r instead of indexed while loading a constant pointer

Author: Richard Plangger Branch: s390x-backend Changeset: r82169:0274e5b9e6d2 Date: 2016-02-12 09:29 +0100 http://bitbucket.org/pypy/pypy/changeset/0274e5b9e6d2/ Log:fixed test. emitting gc_load_r instead of indexed while loading a constant pointer LD has only 12 bit unsigned offset,

[pypy-commit] pypy s390x-backend: replaced arith left shift with logical (other backends do not use arith shift there!),

Author: Richard Plangger Branch: s390x-backend Changeset: r82174:2b50cfaee409 Date: 2016-02-12 12:39 +0100 http://bitbucket.org/pypy/pypy/changeset/2b50cfaee409/ Log:replaced arith left shift with logical (other backends do not use arith shift there!), do not call memcpy for strings a

[pypy-commit] extradoc extradoc: booked train ticket now, added one day for skiing

Author: Richard Plangger Branch: extradoc Changeset: r5606:f23ab745f652 Date: 2016-02-15 09:54 +0100 http://bitbucket.org/pypy/extradoc/changeset/f23ab745f652/ Log:booked train ticket now, added one day for skiing diff --git a/sprintinfo/leysin-winter-2016/people.txt b/sprintinfo/leysin-win

[pypy-commit] pypy s390x-backend: separated two needs in the regalloc, ensure_reg always returns a register (a pool location is never returned), ensure_reg_or_pool now either returns a pool loc or reg

Author: Richard Plangger Branch: s390x-backend Changeset: r82253:13c8dfe611d6 Date: 2016-02-15 10:52 +0100 http://bitbucket.org/pypy/pypy/changeset/13c8dfe611d6/ Log:separated two needs in the regalloc, ensure_reg always returns a register (a pool location is never returned), ensure_r

[pypy-commit] pypy s390x-backend: sync to remote did not work, fixed more calls to ensure_reg

Author: Richard Plangger Branch: s390x-backend Changeset: r82256:4d0391440a29 Date: 2016-02-15 11:48 +0100 http://bitbucket.org/pypy/pypy/changeset/4d0391440a29/ Log:sync to remote did not work, fixed more calls to ensure_reg diff --git a/rpython/jit/backend/zarch/helper/regalloc.py b/rpyth

[pypy-commit] pypy s390x-backend: added te temp_boxes too early, added comments

Author: Richard Plangger Branch: s390x-backend Changeset: r82257:c587ef1d34a4 Date: 2016-02-15 12:07 +0100 http://bitbucket.org/pypy/pypy/changeset/c587ef1d34a4/ Log:added te temp_boxes too early, added comments diff --git a/rpython/jit/backend/zarch/regalloc.py b/rpython/jit/backend/zarch/

[pypy-commit] pypy s390x-backend: stacklet switch removed one register move, rescuing f8-f15

Author: Richard Plangger Branch: s390x-backend Changeset: r82258:2d699125ebd7 Date: 2016-02-15 12:26 +0100 http://bitbucket.org/pypy/pypy/changeset/2d699125ebd7/ Log:stacklet switch removed one register move, rescuing f8-f15 diff --git a/rpython/translator/c/src/stacklet/switch_s390x_gcc.h

[pypy-commit] pypy s390x-backend: merged deafult

Author: Richard Plangger Branch: s390x-backend Changeset: r82259:39811a0085e6 Date: 2016-02-15 12:28 +0100 http://bitbucket.org/pypy/pypy/changeset/39811a0085e6/ Log:merged deafult diff --git a/lib-python/2.7/distutils/command/build_ext.py b/lib-python/2.7/distutils/command/build_ext.py ---

[pypy-commit] pypy s390x-backend: remove _vmprof when translating for s390x (for now)

Author: Richard Plangger Branch: s390x-backend Changeset: r82260:11fdfd9e852f Date: 2016-02-15 13:25 +0100 http://bitbucket.org/pypy/pypy/changeset/11fdfd9e852f/ Log:remove _vmprof when translating for s390x (for now) diff --git a/pypy/config/pypyoption.py b/pypy/config/pypyoption.py --- a/p

[pypy-commit] pypy s390x-backend: extended comment + issue in libc (rposix macro expansion)

Author: Richard Plangger Branch: s390x-backend Changeset: r82261:bb5b1c0fef8d Date: 2016-02-15 13:29 +0100 http://bitbucket.org/pypy/pypy/changeset/bb5b1c0fef8d/ Log:extended comment + issue in libc (rposix macro expansion) diff --git a/rpython/rlib/rposix.py b/rpython/rlib/rposix.py --- a/r

  1   2   3   4   5   6   7   8   9   10   >