summaryrefslogtreecommitdiff
path: root/mpn
Commit message (Collapse)AuthorAgeFilesLines
* Regenerate for:Kevin Ryde2001-11-301-1/+1
| | | | * mpn/Makefile.am (EXTRA_DIST): Add Makeasm.am.
* * mpn/Makefile.am (EXTRA_DIST): Add Makeasm.am.Kevin Ryde2001-11-301-1/+1
|
* Regenerate for:Kevin Ryde2001-11-291-1/+1
| | | | | Amend to: * mpn/Makefile.am (TARG_DIST): Add s390.
* Amend to:Kevin Ryde2001-11-291-1/+1
| | | | * mpn/Makefile.am (TARG_DIST): Add s390.
* * mpn/s370: Rename to s390.Kevin Ryde2001-11-291-0/+0
|
* Regenerate for:Kevin Ryde2001-11-291-1/+1
| | | | * mpn/Makefile.am (TARG_DIST): Add s370.
* * mpn/Makefile.am (TARG_DIST): Add s370.Kevin Ryde2001-11-291-1/+1
|
* Something strange is going on, this file lives in Attictege2001-11-291-29/+0
| | | | but "cvs update" checks it out anyway. Try deleting it again!
* * mpn/s370/gmp-mparam.h: New file.Kevin Ryde2001-11-291-0/+57
| | | | | These values from l005046.zseriespenguins.ihost.com, whatever that machine actually is.
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-291-11/+13
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-291-14/+14
| | | | | The MUL and SQR thresholds don't seem very reproducible, which is not good, will have to look into that.
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-291-15/+17
| | | | | The measuring isn't particularly stable on TOOM3_SQR_THRESHOLD, but this is close enough for now.
* Identify the CPU in the header line.Kevin Ryde2001-11-291-1/+1
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-291-30/+30
| | | | Not up-to-date, but at least in the new style (ie. no #ifndefs).
* Identify the CPU in the header line.Kevin Ryde2001-11-291-1/+1
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-292-63/+47
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-291-9/+11
| | | | | The SQR thresholds looked a bit weird in the measured results, those values left unchanged until it can be investigated.
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-292-21/+21
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-292-25/+26
|
* Identify the CPU in the header line.Kevin Ryde2001-11-291-1/+1
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-293-33/+36
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-294-44/+47
|
* * mpn/*/gmp-mparam.h: Update measured thresholds.Kevin Ryde2001-11-281-50/+33
| | | | | These are not up-to-date, in particular it misses out on powm recent changes, but is better than nothing.
* Fix logic for more_carries scalar loop.tege2001-11-222-8/+10
|
* *** empty log message ***tege2001-11-201-1/+1
|
* *** empty log message ***tege2001-11-201-4/+13
|
* Add a copyright year.Kevin Ryde2001-11-181-1/+1
|
* More of:Kevin Ryde2001-11-151-3/+11
| | | | * mpn/x86/pentium4/README: New file.
* * mpn/x86/t-zdisp2.pl: New file.Kevin Ryde2001-11-151-0/+138
|
* * mpn/x86/pentium4/README: New file.Kevin Ryde2001-11-151-0/+98
|
* Misc changes to make lshift.asm and rshift.asm like.tege2001-11-152-4/+1
|
* Add a copyright year.tege2001-11-1511-14/+14
|
* Amend a comment from:Kevin Ryde2001-11-071-1/+1
| | | | * mpn/x86/mod_34lsub1.asm: New file.
* * mpn/x86/k6/mod_34lsub1.asm, mpn/x86/k6/k62mmx/copyd.asm: Use ZdispKevin Ryde2001-11-072-3/+3
| | | | to force good code alignment.
* * mpn/x86/k7/mmx/lshift.asm, mpn/x86/k7/mmx/rshift.asm: Use Zdisp toKevin Ryde2001-11-072-8/+8
| | | | | | | | force code size for computed jumps. gas new enough to support mmx leaves 0 displacements alone, so this change doesn't fix an actual problem, just ensures we get what we want irrespective of the "optimizations" an assembler might apply.
* * mpn/x86/x86-defs.m4 (Zdisp): More instructions.Kevin Ryde2001-11-071-0/+4
|
* * mpn/x86/pentium/sqr_basecase.asm, mpn/x86/k7/mmx/mod_1.asm,Kevin Ryde2001-11-071-1/+1
| | | | | | mpn/x86/k7/mmx/popham.asm: Remove some unnecessary "0" address offsets. This one was probably a leftover due to starting from the K6 code.
* * mpn/x86/pentium/sqr_basecase.asm, mpn/x86/k7/mmx/mod_1.asm,Kevin Ryde2001-11-071-1/+1
| | | | mpn/x86/k7/mmx/popham.asm: Remove some unnecessary "0" address offsets.
* * mpn/x86/k7/mmx/mod_1.asm, mpn/x86/pentium/sqr_basecase.asm,Kevin Ryde2001-11-071-2/+2
| | | | mpn/x86/k7/mmx/popham.asm: Remove some unnecessary "0" address offsets.
* Work around gas bug.tege2001-11-071-10/+21
|
* (PROLOGUE): Change alignment to 8 (probably a goodtege2001-11-071-1/+1
| | | | idea in general; required for hppa/hppa1_1/udiv_qrnnd.asm).
* * mpn/x86/k7/sqr_basecase.asm: Fix a 0(%edi) to use Zdisp, so theKevin Ryde2001-11-051-2/+2
| | | | | | | computed jumps hit the right spot on old gas. Not a problem on newer gas, which leaves 0(%edi) as written. Looks like a bit of testing or debugging accidentally left in.
* * mpn/lisp/gmpasm-mode.el (comment-start-skip): Correction to the wayKevin Ryde2001-11-041-30/+29
| | | | | | the first \( \) pair is setup. (gmpasm-font-lock-keywords): Don't fontify the space before a "#" etc. Misc tweaks to some comments.
* Fix incorrect header comments.tege2001-11-031-4/+2
|
* * mpn/x86/pentium4/mmx/popham.asm: New file.Kevin Ryde2001-11-021-0/+188
|
* * mpn/x86/x86-defs.m4 (psadbw): New macro.Kevin Ryde2001-11-021-27/+1
| | | | * mpn/x86/k7/mmx/popham.asm: Use it.
* * mpn/x86/x86-defs.m4 (psadbw): New macro.Kevin Ryde2001-11-021-0/+72
|
* * mpn/x86/mod_34lsub1.asm: New file.Kevin Ryde2001-10-311-0/+173
|
* * mpn/x86/k7/mod_34lsub1.asm: New file.Kevin Ryde2001-10-301-0/+179
|
* Save 2 cycles with jg:Kevin Ryde2001-10-301-11/+8
| | | | * mpn/x86/k6/mod_34lsub1.asm: New file.
View Lorry Controller Status