Login
Username:

Password:

Remember me



Lost Password?

Register now!
Main Menu
Who's Online
2 user(s) are online (1 user(s) are browsing Forum)

Members: 1
Guests: 1

wawa, more...

Browsing this Thread:   1 Anonymous Users



« 1 ... 4 5 6 (7) 8 9 10 ... 26 »


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
so here is an example of the overloaded functions i mentioned:

Quote:

[ 2%] Building CXX object Source/JavaScriptCore/CMakeFiles/LLIntOffsetsExtractor.dir/llint/LLIntOffsetsExtractor.cpp.obj
In file included from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/Structure.h:36:0,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/ButterflyInlines.h:34,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/JSArray.h:25,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/ArrayProfile.h:30,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/llint/LLIntOffsetsExtractor.cpp:28:
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/PropertyNameArray.h: In member function 'void JSC::PropertyNameArray::add(uint32_t)':
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/PropertyNameArray.h:67:41: error: call of overloaded 'from(JSC::VM*&, uint32_t&)' is ambiguous
add(Identifier::from(m_vm, index));
^
In file included from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/PropertyName.h:29:0,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/Structure.h:35,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/ButterflyInlines.h:34,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/JSArray.h:25,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/ArrayProfile.h:30,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/llint/LLIntOffsetsExtractor.cpp:28:
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/Identifier.h:136:23: note: candidate: static JSC::Identifier JSC::Identifier::from(JSC::VM*, unsigned int)
static Identifier from(VM*, unsigned y);
^~~~
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/Identifier.h:137:23: note: candidate: static JSC::Identifier JSC::Identifier::from(JSC::VM*, int)
static Identifier from(VM*, int y);
^~~~
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/Identifier.h:138:23: note: candidate: static JSC::Identifier JSC::Identifier::from(JSC::VM*, double)
static Identifier from(VM*, double y);
^~~~
In file included from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/CodeBlock.h:62:0,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/llint/LLIntOffsetsExtractor.cpp:29:
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/LazyOperandValueProfile.h: In member function 'unsigned int JSC::LazyOperandValueProfileKey::hash() const':
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/LazyOperandValueProfile.h:74:45: error: call of overloaded 'intHash(const unsigned int&)' is ambiguous
return WTF::intHash(m_bytecodeOffset) + m_operand.offset();
^
In file included from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashTraits.h:24:0,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashTable.h:32,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashSet.h:27,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/heap/CodeBlockSet.h:31,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/heap/Heap.h:26,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/heap/DeferGC.h:29,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/runtime/ConcurrentJITLock.h:29,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/bytecode/ArrayProfile.h:29,
from /home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/JavaScriptCore/llint/LLIntOffsetsExtractor.cpp:28:
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashFunctions.h:39:21: note: candidate: unsigned int WTF::intHash(uint8_t)
inline unsigned intHash(uint8_t key8)
^~~~~~~
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashFunctions.h:52:21: note: candidate: unsigned int WTF::intHash(uint16_t)
inline unsigned intHash(uint16_t key16)
^~~~~~~
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashFunctions.h:65:21: note: candidate: unsigned int WTF::intHash(uint32_t)
inline unsigned intHash(uint32_t key)
^~~~~~~
/home/wawa/AROS-source/AROS/local/OdysseyWebBrowser/Source/WTF/wtf/HashFunctions.h:77:21: note: candidate: unsigned int WTF::intHash(uint64_t)
inline unsigned intHash(uint64_t key)
^~~~~~~
Source/JavaScriptCore/CMakeFiles/LLIntOffsetsExtractor.dir/build.make:97: recipe for target 'Source/JavaScriptCore/CMakeFiles/LLIntOffsetsExtractor.dir/llint/LLIntOffsetsExtractor.cpp.obj' failed

the question is what to do here.
perhaps type definitions cuold be redefined/overriden to what osyssey expects, if this is the correct solution that is. still i dont know yet how, and im still researching the icu issue.

Posted on: 3/10 16:11
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2005/9/3 1:24
From Colorado, USA
Group:
Member
Posts: 750
Offline
It looks to me like it wants sized integers which would be an easy fix. If it works on x86, AMD64, and ARM then endianness is a likely issue and isn't as easy.

Posted on: 3/10 18:01
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
Quote:

SamuraiCrow wrote:
It looks to me like it wants sized integers which would be an easy fix.


if it is easy then tell me how to do it by all means. (properly)

Quote:
If it works on x86, AMD64, and ARM then endianness is a likely issue and isn't as easy.


dont forget odyssey comes from a big endian system and it works on both os4 and morphos. im also compiling it for aros ppc in parallel, currently stuck at icu as well.

btw i dont think it works on arm yet, the issue being again icu, which wont cross compile.

Posted on: 3/11 4:02
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2005/9/3 1:24
From Colorado, USA
Group:
Member
Posts: 750
Offline
@wawa

The unsigned int type is ambiguous until you know that it's running on a 32-bit processor. That means that whatever is typedef rendered as unsigned int should be uint_32 instead. If it's running on a 64-bit CPU it should be uint_64.

Posted on: 3/11 10:14
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
soo.. for those interested here is a new experimental version compiled with the current icu 60.2. this version unfortunatelly has a bigger executable and may as well need more stack and memory. i have not tested it very well, as i dont have internet well set up for aros hosted.
the old executable is stll included for comparison.

https://1drv.ms/u/s!AuvzEgbTrkxPgYQKmFIXirmLMO3R3Q

Posted on: 3/21 3:24
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2006/4/7 1:34
Group:
Member
Posts: 658
Offline
@wawa

Exzellent! Could you please tell us, which AROS version we should or have to use for testing?

Posted on: 3/21 3:44
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
im currently still compiling it for x86 abiv1.

Posted on: 3/21 4:32
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
for reference compiling icu for m68k we have similar problem of ambigious overloads as odyssey/webkit itself:
/home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu/source/i18n/compactdecimalformat.cppIn function 'int32_t icu_60::populatePrefixSuffix(const char*, int32_t, const icu_60::UnicodeString&, UHashtable*, UBool, UErrorCode&)':
/
home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu/source/i18n/compactdecimalformat.cpp:777:45errorcall of overloaded 'indexOf(char, int)' is ambiguous
   int32_t semiPos 
formatStr.indexOf(';'0);
                                             ^
In file included from /home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu/source/common/charstr.h:18:0,
                 
from /home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu/source/i18n/compactdecimalformat.cpp:17:
/
home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu/source/common/unicode/unistr.h:4282:1notecandidateint32_t icu_60::UnicodeString::indexOf(char16_tint32_t) const
 
UnicodeString::indexOf(char16_t c,
 ^~~~~~~~~~~~~
/
home/wawa/aros-m68k-630/bin/amiga-m68k/Ports/icu4c/icu

i think as soon as i know how to solve these it may occur that icu/odyssey for other platforms as i386 are compilable.

Posted on: 3/21 6:39

Edited by wawa on 2018/3/22 0:17:04
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2007/4/16 6:26
From Norway
Group:
Member
Posts: 4010
Offline
That is fantasic news :)
Thank U. Will test later today.

Edit: tested and can not say I notice any difference. Is there anything I could test?

Posted on: 3/21 6:50

Edited by nikolaos on 2018/3/21 12:28:17
_________________
www.aspireos.com
Transfer the post to other applications Transfer


Re: Odyssey web browser future

Joined:
2010/8/30 7:20
Group:
Member
Posts: 1194
Offline
for me its aleady quite reassuring, since it suggests that i have not broken anything significant porting the new version. to be honest i dont know what its advantages are. for me it was about portability and platform independance, which i have been confirmed with to certain extent. unfortunatelly not quite fully, which kind of sets the whole effort a bit under a question mark.

namely, last night i almost cross compiled icu-60.2 for ppc. but then on the last stright, there comes the blow:

genccodecurrently only same-endianness ELF formats are supported
Makefile
:193recipe for target 'packagedata' failed
make
[3]: *** [packagedataError 16
Makefile
:149recipe for target 'all-recursive' failed
make
[2]: *** [all-recursiveError 2
mmakefile
:409recipe for target '/home/wawa/aros-ppc-630/bin/sam440-ppc/gen/contrib/development/libs/icu/icu4c/.installed' failed
make
[1]: *** [/home/wawa/aros-ppc-630/bin/sam440-ppc/gen/contrib/development/libs/icu/icu4c/.installedError 2
[MMAKEmake --no-print-directory TOP=/home/wawa/aros-ppc-630 SRCDIR=/home/wawa/AROS-source/AROS CURDIR=contrib/development/libs/icu TARGET=contrib-icu4c-complete--build_and_install-quick --file=mmakefile contrib-icu4c-complete--build_and_install-quick failed512
[MMAKEErrorError while running make in contrib/development/libs/icuNo such file or directory
Makefile
:237recipe for target 'contrib-icu4c' failed
make
: *** [contrib-icu4cError 10
wawa
@wawa-virtual-machine:~/aros-ppc-630$


which issue still stands as of today (source/tools/toolutil/pkg_genc.cpp):
747        *pIsBigEndian=(UBool)(buffer.header32.e_ident[EI_DATA]==ELFDATA2MSB);
748        if(*pIsBigEndian!=U_IS_BIG_ENDIAN) {
749            fprintf(stderr"genccode: currently only same-endianness ELF formats are supportedn");
750            exit(U_UNSUPPORTED_ERROR);
751        }
752        /* TODO: Support byte swapping */


and i dont expect that it is a priority fo them. :( ill try to enquire about this. not hoping for much, especially that neither aros ppc platform isnt ready for actualy testing nor i do have hardware in question.

the other possibility would be to actally implement byte swap building these elf binaries, then again i dont see who might care for it.

edit: gosh, these code boxes are annoying, either you have code unreadable using "quote" or you get layout borked..

Posted on: 3/22 0:15

Edited by wawa on 2018/3/22 0:47:42
Edited by wawa on 2018/3/22 0:55:36
Transfer the post to other applications Transfer



« 1 ... 4 5 6 (7) 8 9 10 ... 26 »



You can view topic.
You cannot start a new topic.
You cannot reply to posts.
You cannot edit your posts.
You cannot delete your posts.
You cannot add new polls.
You cannot vote in polls.
You cannot attach files to posts.
You cannot post without approval.

[Advanced Search]


Search
Top Posters
1 paolone
paolone
4427
2 magorium
magorium
4095
3 nikolaos
nikolaos
4010
4 phoenixkonsole
phoenixkonsole
3920
5 deadwood
deadwood
2917
6 ncafferkey
ncafferkey
2794
7 mazze
mazze
2221
8 clusteruk
clusteruk
2112
9 damocles
damocles
1789
10 BSzili
BSzili
1511
© 2004-2018 AROS Exec