What You Can Help

64 views
Skip to first unread message

wbrana

unread,
Aug 22, 2012, 9:29:18 AM8/22/12
to x32...@googlegroups.com
Section "What You Can Help" could be extended with:
Persuade following companies to support X32 ABI in their software.
X32 ABI in Linux won't be useful for most users if these companies
won't support it.

Microsoft in Windows and Visual Studio (most important)
- http://mymfe.microsoft.com/Windows%207/Feedback.aspx?formID=195
- http://wfp.microsoft.com/
NVIDIA in drivers
Google in Chrome
Mozilla in Firefox
Oracle in Java and VirtualBox

William Brana

unread,
Aug 22, 2012, 11:42:23 AM8/22/12
to x32...@googlegroups.com

Daniel Schepler

unread,
Aug 22, 2012, 12:45:30 PM8/22/12
to x32...@googlegroups.com
On Wednesday, August 22, 2012 6:29:18 AM UTC-7, William Brana wrote:
Microsoft in Windows and Visual Studio (most important)
- http://mymfe.microsoft.com/Windows%207/Feedback.aspx?formID=195
- http://wfp.microsoft.com/

I don't understand.  What possible difference could it make to us whether Windows supports an X32 style ABI?  (Other than that wine would have to be updated if they did add such a thing.)

On the other things you listed, I'd agree.
--
Daniel Schepler

William Brana

unread,
Aug 22, 2012, 12:48:24 PM8/22/12
to x32...@googlegroups.com
It is more likely that NVIDIA, Google, Mozilla, Oracle will support X32 on Linux if Microsoft will support X32 on Windows.

Mike Frysinger

unread,
Aug 22, 2012, 2:27:21 PM8/22/12
to x32...@googlegroups.com
On Wed, Aug 22, 2012 at 9:48 AM, William Brana <wbr...@gmail.com> wrote:
> It is more likely that NVIDIA, Google, Mozilla, Oracle will support X32 on
> Linux if Microsoft will support X32 on Windows.

i don't think that's true. it makes no sense for microsoft to do x32,
nor does it have any bearing whatsoever on linux ports of nvidia
drivers or google's code base or mozilla's code base. i'm sure the
browser venders will eventually make it over (they're open source
projects after all), but it would be good to lobby nvidia to make x32
versions of their binary drivers.
-mike

William Brana

unread,
Aug 22, 2012, 2:57:59 PM8/22/12
to x32...@googlegroups.com
NVIDIA shares code between Windows and Linux.
It is unlikely there will be X32 Linux driver if there won't be X32 Windows driver.

Mike Frysinger

unread,
Aug 22, 2012, 3:11:28 PM8/22/12
to x32...@googlegroups.com
do you actually know how the x64 code in windows works ? it isn't
like the 64bit abi in linux. so by your logic, i have no idea why
nvidia did a port to linux's 64bit abi because microsoft certainly
doesn't support it.

if anything, windows x64 is more like x32 -- sizeof(long) == 4.
http://msdn.microsoft.com/en-us/library/7kcdt6fy.aspx
-mike

William Brana

unread,
Aug 22, 2012, 3:22:11 PM8/22/12
to x32...@googlegroups.com
I have never used 64-bit Windows.
Size of long isn't important, but register count/size and pointer size, which is same for Windows and Linux.

Mike Frysinger

unread,
Aug 22, 2012, 3:39:59 PM8/22/12
to x32...@googlegroups.com
and register/count size is exactly the same. only the pointer size is
different, but even that is irrelevant. your logic makes no sense --
nvidia isn't going to be held up doing a port to Linux/x32 by
*anything* microsoft does (or does not do as is the case here).

if you want to see an nvidia/x32 port, then go post to nvidia's linux
forums asking for it.
-mike

William Brana

unread,
Aug 22, 2012, 3:44:50 PM8/22/12
to x32...@googlegroups.com
I already asked nvidia at their forum. There was no response whether it will be supported or not.

Gianni.

unread,
Aug 22, 2012, 3:48:16 PM8/22/12
to x32...@googlegroups.com
I'm sorry, but what does a *driver* has to do to support x32? After all,
kernel-space is still x86-64, no?

Michael LIAO

unread,
Aug 22, 2012, 4:10:59 PM8/22/12
to x32...@googlegroups.com
It should refer to the user-space GL driver.

- Michael

William Brana

unread,
Aug 22, 2012, 4:26:30 PM8/22/12
to x32...@googlegroups.com

Mike Frysinger

unread,
Aug 22, 2012, 5:31:37 PM8/22/12
to x32...@googlegroups.com
On Wed, Aug 22, 2012 at 3:48 PM, Gianni. wrote:
> On Wednesday 22 August 2012 15:39:59 Mike Frysinger wrote:
>> On Wed, Aug 22, 2012 at 3:22 PM, William Brana wrote:
>> > I have never used 64-bit Windows.
>> > Size of long isn't important, but register count/size and pointer size,
>> > which is same for Windows and Linux.
>>
>> and register/count size is exactly the same. only the pointer size is
>> different, but even that is irrelevant. your logic makes no sense --
>> nvidia isn't going to be held up doing a port to Linux/x32 by
>> *anything* microsoft does (or does not do as is the case here).
>>
>> if you want to see an nvidia/x32 port, then go post to nvidia's linux
>> forums asking for it.
>
> I'm sorry, but what does a *driver* has to do to support x32? After all,
> kernel-space is still x86-64, no?

binary graphics drivers usually have two components -- the kernel and
userspace. you're correct the kernel part needs no changes at all,
but the userland part does. specifically, nvidia's driver package
includes (among other things):
libcuda.so.1
libOpenCL.so.1
libglx.so.1
libGL.so.1
it ships both x86 (32bit) and x86_64 (64bit) libraries so that the
driver works with either userland. without these pieces, you can't
get accelerated hardware graphics with an x32 userland :(.

nvidia isn't unique in this setup -- pretty much everyone doing closed
binary graphics drivers splits it this way. the arm community went
through a painful transition when they changed the userland ABI from
soft floating point (emulating in software) to hard floating point
(using hardware instructions).
-mike

William Brana

unread,
Aug 22, 2012, 6:29:50 PM8/22/12
to x32...@googlegroups.com
open source drivers are split in 4 parts:
1. kernel
userspace:
2. libdrm
3. xf86-video-intel, xf86-video-ati
4. Mesa
Reply all
Reply to author
Forward
0 new messages