Re: [patch] add x32 support
On Monday 14 January 2013 14:36:43 Torbjorn Granlund wrote: > At some point, we'd like to make the assembly code in GMP support x32. how so ? for the most part, existing x86_64 assembly should "just work" for x32. pointers tend to be where things get into trouble, but otherwise x32 is simply x86_64 with different sizeof(long) & sizeof(void*). 64bit types work just fine and with the same performance as in x86_64 -- no ugliness like 32bit x86. `make check` passes for me w/gmp in x32. so at least something is working. -mike signature.asc Description: This is a digitally signed message part. ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
Richard Henderson writes: On 01/10/2013 12:46 PM, Torbjorn Granlund wrote: > (I wonder if this can become a robust ABI with current hardware, unless > we pay a major performance price. It seems like address arithmetic by > means of hardware indexing, lea, etc, will be 64-bits, while we want to > compute mod 2^32.) The addr32 prefix takes care of this. Can you please explain a bit about how that works? Using insn(%rX), reg gives me no addr32 prefix, but insn(%eX), reg does indeed. Is that register syntax change the standard way of handling this? For assembly code that needs to support both the 64-bit ABI ans x32, it might be more convenient to conditionally put an explicit ".byte 0x67" at the preceding line. Apparently gas also accept "addr32" but then prints "Warning: stand-alone `addr32' prefix". An optimisation is to use a 64-bit expresson if one knows that the expresson >= 0 and < 2^32. Correct? So that saves a byte. Do you know if there are other costs, i.e., decoding or execution costs associated to addr32? (OK, insn aligment vs cache lines matters, and might cause addr32 to either slow down or speed up exeuction.) At some point, we'd like to make the assembly code in GMP support x32. I think it should already be fine, without any addr32 prefixing, but I am not sure. For most reasonable indexing, I'd say that one should not magically end up with a non-canonical address. Do you agree? -- Torbjörn ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
On 01/10/2013 12:46 PM, Torbjorn Granlund wrote: (I wonder if this can become a robust ABI with current hardware, unless we pay a major performance price. It seems like address arithmetic by means of hardware indexing, lea, etc, will be 64-bits, while we want to compute mod 2^32.) The addr32 prefix takes care of this. r~ ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
On Thursday 10 January 2013 15:46:55 Torbjorn Granlund wrote: > I was mainly concerned about your statement that "the x32 ABI slipped > through the cracks". I cannot find any prior communications about this. i was referring to the people who wrote the first patch which i've seen rebased/semi-rewritten rather than the gmp peeps missing it. i don't think it has been submitted yet. > How could I test this? do you have an x86_64 processor and running linux-3.4+ ? the fastest answer is to enable x32 support in your kernel config (CONFIG_X86_X32) and then use a Gentoo chroot: http://distfiles.gentoo.org/releases/amd64/autobuilds/current-stage3/ unpack that into a dir, chroot into it, then you'll have a multilib setup with x32 as the default ABI. if you have any troubles, feel free to ask and i'll see what's what. -mike signature.asc Description: This is a digitally signed message part. ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
On Thu, 10 Jan 2013, Marc Glisse wrote: On Thu, 10 Jan 2013, Torbjorn Granlund wrote: How could I test this? I don't remember what linux you have access to. Debian for instance has packages in "experimental" named libx32gcc1, libc6-x32, etc. Then you just need to compile with gcc -mx32. Hmm, it also requires installing a kernel compiled with a different option, etc, so scratch that, sorry :-( -- Marc Glisse ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
On Thu, 10 Jan 2013, Torbjorn Granlund wrote: How could I test this? I don't remember what linux you have access to. Debian for instance has packages in "experimental" named libx32gcc1, libc6-x32, etc. Then you just need to compile with gcc -mx32. -- Marc Glisse ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
Mike Frysinger writes: On Thursday 10 January 2013 15:05:12 Torbjorn Granlund wrote: > Mike Frysinger writes: > > looks like support for the x32 ABI slipped through the cracks. here's > the patch updated to latest hg branch (http://gmplib.org:8000/gmp). > > What is this? Please give some context. a simple google for "x32 abi": http://en.wikipedia.org/wiki/X32_ABI https://sites.google.com/site/x32abi/ Right another ABI. I was mainly concerned about your statement that "the x32 ABI slipped through the cracks". I cannot find any prior communications about this. How could I test this? (I wonder if this can become a robust ABI with current hardware, unless we pay a major performance price. It seems like address arithmetic by means of hardware indexing, lea, etc, will be 64-bits, while we want to compute mod 2^32.) -- Torbjörn ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
On Thursday 10 January 2013 15:05:12 Torbjorn Granlund wrote: > Mike Frysinger writes: > > looks like support for the x32 ABI slipped through the cracks. here's > the patch updated to latest hg branch (http://gmplib.org:8000/gmp). > > What is this? Please give some context. a simple google for "x32 abi": http://en.wikipedia.org/wiki/X32_ABI https://sites.google.com/site/x32abi/ -mike signature.asc Description: This is a digitally signed message part. ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
Re: [patch] add x32 support
Mike Frysinger writes: looks like support for the x32 ABI slipped through the cracks. here's the patch updated to latest hg branch (http://gmplib.org:8000/gmp). What is this? Please give some context. -- Torbjörn ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
[patch] add x32 support
looks like support for the x32 ABI slipped through the cracks. here's the patch updated to latest hg branch (http://gmplib.org:8000/gmp). -mike diff -r f47abbae9591 configure.ac --- a/configure.ac Thu Jan 10 13:37:36 2013 +0100 +++ b/configure.ac Thu Jan 10 14:53:40 2013 -0500 @@ -1668,9 +1668,21 @@ CALLING_CONVENTIONS_OBJS_64='amd64call.lo amd64check$U.lo' SPEED_CYCLECOUNTER_OBJ_64=x86_64.lo cyclecounter_size_64=2 - abilist="64 32" + + cclist_x32="gcc" + gcc_x32_cflags="$gcc_cflags -mx32" + gcc_x32_cflags_optlist="$gcc_64_cflags_optlist" + CALLING_CONVENTIONS_OBJS_x32="$CALLING_CONVENTIONS_OBJS_64" + SPEED_CYCLECOUNTER_OBJ_x32="$SPEED_CYCLECOUNTER_OBJ_64" + cyclecounter_size_x32="$cyclecounter_size_64" + path_x32="$path_64" + limb_x32=longlong + any_x32_testlist="sizeof-long-4" + + abilist="64 x32 32" if test "$enable_assembly" = "yes" ; then extra_functions_64="invert_limb_table" + extra_functions_x32=$extra_functions_64 fi case $host in @@ -3427,7 +3439,7 @@ GMP_INCLUDE_MPN(x86/darwin.m4) ;; esac ;; -64) +64|x32) GMP_INCLUDE_MPN(x86_64/x86_64-defs.m4) AC_DEFINE(HAVE_HOST_CPU_FAMILY_x86_64) case $host in signature.asc Description: This is a digitally signed message part. ___ gmp-devel mailing list gmp-devel@gmplib.org http://gmplib.org/mailman/listinfo/gmp-devel
