As far as I can see from the compiler errors the patch made no difference.
Line 484 in jsregexp.cc is
IrregexpResult res = RegExpImpl::IrregexpExecOnce(
jsregexp, subject, previous_index, Vector<int32_t>(registers.vector(),
registers.length()));
As the patch should make registers.vector() return int32_t* I cannot
understand how the error
rc/jsregexp.cc:484: error: no matching function for call to
'v8::internal::Vector<long int>::Vector(int*, int)'
can occur. Could you please check once more how the template argument
(int32_t) can be different from the type of first argument to the Vector
constructor. I might have missed something in the patch, as I am noy able to
reproduce the problem on Liunx.
Regards,
Søren
On Wed, Aug 25, 2010 at 00:05, pikpik <[email protected]> wrote:
> Hi,
>
> On Aug 24, 4:21 am, Søren Gjesse <[email protected]> wrote:
> > Hi,
> >
> > Looking at the compiler once more it looks as if int32_t is long on
> MINIX,
> > whereas on Linux it is int. In
> https://gforge.cs.vu.nl/gf/project/minix/scmsvn/?action=browse&path=/...
> > i32_t is typedef to long when __SIZEOF_LONG__ == 4.
> >
> > I found some more mixed uses int/int32_t, which might be causing the
> > compiler issues you are facing. Could you maybe try the patch below,
> which
> > should get rid of more of these, and see where that brings you.
>
> Yes, I don't have much experience, but I have noticed a few
> differences (this and "int64_t," for which I'm using a patch to the
> "ansi.h" header file) between MINIX and other systems. Is MINIX's
> difference too much to be supported easily (perhaps changes should be
> done to it instead of V8?), or is it a lot of work like any system?
>
> > Regards,
> > Søren
>
> I updated to revision 5322 and applied the patch (it appears to be
> applied correctly).
>
> Here's the last part of my build:
>
> /usr/gnu/bin/g++ -o obj/release/jsregexp.o -c -Wall -Werror -W -Wno-
> unused-parameter -Wnon-virtual-dtor -pedantic -m32 -O3 -fomit-frame-
> pointer -ansi -Wno-char-subscripts -fno-strict-aliasing -fno-rtti -fno-
> exceptions -Wall -Werror -W -Wno-unused-parameter -Wnon-virtual-dtor -
> pedantic -m32 -O3 -fomit-frame-pointer -ansi -Wno-char-subscripts -fno-
> strict-aliasing -DV8_TARGET_ARCH_IA32 -DENABLE_VMSTATE_TRACKING -
> DENABLE_LOGGING_AND_PROFILING -DENABLE_DEBUGGER_SUPPORT -Isrc src/
> jsregexp.cc
>
> src/jsregexp.cc: In static member function 'static
> v8::internal::RegExpImpl::IrregexpResult
>
> v8::internal::RegExpImpl::IrregexpExecOnce(v8::internal::Handle<v8::internal::JSRegExp>,
> v8::internal::Handle<v8::internal::String>, int,
> v8::internal::Vector<long int>)':
>
> src/jsregexp.cc:410: error: no matching function for call to
>
> 'v8::internal::NativeRegExpMacroAssembler::Match(v8::internal::Handle<v8::internal::Code>&,
> v8::internal::Handle<v8::internal::String>&, long int*, int, int&)'
>
> src/regexp-macro-assembler.h:186: note: candidates are: static
> v8::internal::NativeRegExpMacroAssembler::Result
>
> v8::internal::NativeRegExpMacroAssembler::Match(v8::internal::Handle<v8::internal::Code>,
> v8::internal::Handle<v8::internal::String>, int*, int, int) <near
> match>
>
> src/jsregexp.cc: In static member function 'static
> v8::internal::Handle<v8::internal::Object>
>
> v8::internal::RegExpImpl::IrregexpExec(v8::internal::Handle<v8::internal::JSRegExp>,
> v8::internal::Handle<v8::internal::String>, int,
> v8::internal::Handle<v8::internal::JSArray>)':
>
> src/jsregexp.cc:484: error: no matching function for call to
> 'v8::internal::Vector<long int>::Vector(int*, int)'
>
> src/utils.h:318: note: candidates are:
> v8::internal::Vector<T>::Vector(T*, int) [with T = long int] <near
> match>
>
> src/utils.h:317: note:
> v8::internal::Vector<T>::Vector() [with T = long int]
>
> src/utils.h:315: note: v8::internal::Vector<long
> int>::Vector(const v8::internal::Vector<long int>&)
>
> scons: *** [obj/release/jsregexp.o] Error 1
>
> scons: building terminated because of errors.
>
>
> I guess that this is caused by MINIX's differing definition of
> "int32_t" and might lead to changes in related parts of V8? I'm sorry
> if this is an abnormally large amount of work.
>
> Thanks,
> pikpik
>
> --
> v8-users mailing list
> [email protected]
> http://groups.google.com/group/v8-users
>
--
v8-users mailing list
[email protected]
http://groups.google.com/group/v8-users
