harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From xiaoming gu <xiaoming...@gmail.com>
Subject Re: discussion for H5022
Date Fri, 06 Feb 2009 07:00:58 GMT
I read the code about heap base address in Ia32InstCodeSelector.cpp and
checked the operands created by those parts of code with a simple test case
on a Windows64 machine. The code itself and the practical results both show
all related operands created in HIR2LIR pass are in 64-bit size. Thanks.

Xiaoming

On Fri, Feb 6, 2009 at 10:17 AM, Xiao-Feng Li <xiaofeng.li@gmail.com> wrote:

> Does newImmOpnd() generate 64bit imm?
>
> Thanks,
> xiaofeng
>
> On Thu, Feb 5, 2009 at 4:20 PM, xiaoming gu <xiaoming.gu@gmail.com> wrote:
> > I did some tests on 64-bit Windows and found sizeof(POINTER_SIZE_INT)=8.
> So
> > I think no truncation happens. And I could NOT find the code mentioned in
> > [1] and maybe it is because of some rollback. What's your opinion?
> >
> > Xiaoming
> >
> > [1] - https://issues.apache.org/jira/browse/HARMONY-5022
> >
> > On Mon, Nov 3, 2008 at 2:00 PM, xiaoming gu <xiaoming.gu@gmail.com>
> wrote:
> >
> >> So in 64-bit platform with compressed pointer, the base address for heap
> is
> >> still 64-bit but the offset is 32-bit.
> >>
> >> Got it. Thanks. -Xiaoming
> >>
> >>
> >> On Mon, Nov 3, 2008 at 12:21 PM, Xiao-Feng Li <xiaofeng.li@gmail.com
> >wrote:
> >>
> >>> Xiaoming, I think this work is only related with X86-64, the 64bit
> >>> machine. In a 32bit platform, there is no such concept of 64-bit base
> >>> address. In 64-bit platform with compressed pointer, we have a base
> >>> address and a 32-bit offset for one pointer.
> >>>
> >>> Thanks,
> >>> xiaofeng
> >>>
> >>> On Mon, Nov 3, 2008 at 11:52 AM, xiaoming gu <xiaoming.gu@gmail.com>
> >>> wrote:
> >>> > Hi, guys. I'm studying H5022 - incorrect codegeneration of compressed
> >>> obj
> >>> > arithmetics[1]. In my understanding, the problem is that current JIT
> >>> casts
> >>> > heap base address from 64-bit to 32-bit compulsorily in IA32 but that
> >>> > address must be in 64-bit even in a 32-bit platform. If I catch the
> >>> point of
> >>> > this jira correctly, I'll try to delete the casts in
> >>> > Ia32InstCodeSelector.cpp recently. Any comment? Thanks.
> >>> >
> >>> > Xiaoming
> >>> >
> >>> > [1] - https://issues.apache.org/jira/browse/HARMONY-5022
> >>> >
> >>>
> >>>
> >>>
> >>> --
> >>> http://xiao-feng.blogspot.com
> >>>
> >>
> >>
> >
> >
> > --
> > China Runtime Technologies Lab,
> > SSG/SSD/MRTC, Intel
> >
>
>
>
> --
> Managed Runtime Technology Center, Intel
>



-- 
China Runtime Technologies Lab,
SSG/SSD/MRTC, Intel

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message