sbcl Log


Commit Date  
[04d819] by Christophe Rhodes Christophe Rhodes

0.9.8.27:
Merge patch from James Bielman fixing self-build under Win32.
... a little bit of an accident ensued; I ran
canonicalize-whitespace on .sh files, which seemed to
work (except on wc.sh) until it scribbled all over
make.sh, confusing the shell interpreter utterly.
... the resulting files have built the system, nevertheless.

2006-01-09 13:00:17 Tree
[732ad8] by Christophe Rhodes Christophe Rhodes

0.9.8.26:
Merge patch from James Bielman fixing use of GNAME on
fast_bzero* assembly functions.

2006-01-09 12:45:47 Tree
[bb8a05] by Christophe Rhodes Christophe Rhodes

0.9.8.25:
Merge patch from Luis Oliviera (sbcl-devel 2006-01-08) fixing
probe-file on Win32.

2006-01-08 22:39:03 Tree
[9a97f2] by Juho Snellman Juho Snellman

0.9.8.24:
It's x86-64 sign-extension time, once again. The
SB!VM:SIGN-EXTEND VOP wasn't getting selected, causing a full
call for every use. Luckily it's only used only for alien
calls whose return value is declared as a signed byte of 32 or
fewer bits.

2006-01-08 06:31:23 Tree
[e36329] by Juho Snellman Juho Snellman

0.9.8.23:
Add x86-64 support for passing alien callback parameters on the stack.

2006-01-08 03:06:11 Tree
[3a4790] by Juho Snellman Juho Snellman

0.9.8.22:
Some x86-64 SAP-REF setter VOPs for floats were trying to use
the MOVE macro on float registers, causing compile errors in
rare circumstances. Extend x86-64 MOVE to also handle
SINGLE-REG and DOUBLE-REG SCs.

2006-01-08 02:01:42 Tree
[b75e04] by Juho Snellman Juho Snellman

0.9.8.21:
Oops.

2006-01-08 00:59:34 Tree
[6e6670] by Juho Snellman Juho Snellman

0.9.8.20:
Final batch from sbcl-devel "Changes to GENCGC memory zeroing"
in 2005-12). Use hand-coded assembly for zeroing memory in
GENCGC instead of the platform memset/bzero.

* Use MOVNTDQ on x86-64
* Use MOVNTDQ on x86 that supports SSE2 (basically Pentium 4 and
newer)
* Difference to the version posted on sbcl-devel: Do *not* use
the MMX MOVNTQ for x86 that supports MMX but not
SSE2. MOVNTQ apparently had very bad performance on K7
Athlons/Durons.
* Use REP STOSL on remaining x86.

2006-01-08 00:33:13 Tree
[974231] by Juho Snellman Juho Snellman

0.9.8.19:
Changes to GENCGC memory zeroing behaviour that give a big
performance boost, especially on modern processors.

* Instead of zeroing memory by remapping memory with
munmap/mmap at GC time, pages are just marked as needing
zeroing and zeroed with memset when they're added to a new
allocation region. This reduces GC latency both for the
common and worst cases.

* To keep the memory footprint down, clear the pages by
remapping after major GCs (arbitrarily defined as a
collection of generation 2 or older). The memory freed from
a minor GC is just going to get used again immediately , so
releasing them back to the OS would make little sense.

* Add a GENCGC mode (#define READ_PROTECT_FREE_PAGES) for
catching attempts to read unallocated pages

* See sbcl-devel "Changes to GENCGC memory zeroing" in 2005-12
for more details and performance measurements. (Note
that many parts of this patch have already been committed
piecemeal over the last month, this is just the most
significant chunk).

* Performance effect on BSDs (which used a different zeroing
strategy than Linux before this) is unknown.

2006-01-07 18:53:39 Tree
[fd225c] by Juho Snellman Juho Snellman

0.9.8.18:
Rearrange the GENCGC "struct page" a bit to for a more compact
memory representation. Saves memory (about 15MB on x86-64
where the page table is large, a couple of MB on x86). Also
a minor performance improvement thanks to cache issues.

TODO: The size could be still improved by another 15MB on x86-64
by defining the ill-named first_object_offset as an int
instead of long (4 bytes less data and 4 bytes less of padding).
The naive implementation would then limit the maximum region size
to 4GB. Since some low bits in the field are guaranteed to be
zero, a smart implementation could do some shifts and store even
more data. It remains to be seen whether this would be worthwhile.

2006-01-07 18:22:58 Tree
[8c685e] by Christophe Rhodes Christophe Rhodes

0.9.8.17:
Fix a bug in ENSURE-DIRECTORIES-EXIST: merge in
*DEFAULT-PATHNAME-DEFAULTS*.
... this fix may also include a mostly-working set of pathname
functions for Win32. Or it may not. You have been
warned.

2006-01-06 16:44:59 Tree
[aeceaa] by Juho Snellman Juho Snellman

0.9.8.16:
Merge sbcl-devel "Some assembler improvements for x86-64" by
Lutz Euler on 2006-01-05.

Quote from the email:

1. a reduction in core size of 190 KB due to shorter encodings
for common forms of the MOV instruction, and
2. for robustness, better checking of dword-sized immediate
arguments in the assembler with respect to their implicit
sign-extension.

2006-01-06 03:31:26 Tree
[f2d878] by Juho Snellman Juho Snellman

0.9.8.15:
More with-testage. Merge sbcl-devel "[PATCH] callback tests"
by Cyrus Harmon on 2006-01-06.

2006-01-06 02:37:06 Tree
[93be00] by Juho Snellman Juho Snellman

0.9.8.14:
Fix saving a core with callbacks on x86 and x86-64, as
discussed on sbcl-devel "CFFI Callbacks on SBCL" on
2005-12-31. Essentially the problem is that the address of
#'ENTER-ALIEN-CALLBACK is hard-coded into the assembly callback
wrappers, and the address of the function can change when
saving a non-purified core.

* Define a static symbol that contains #'ENTER-ALIEN-CALLBACK
in the value slot.
* Change the x86 / x86-64 wrappers to indirect through the
slot.
* Add minimal test case
* Add a make-config.sh-detected :alien-callbacks feature

2006-01-06 01:11:07 Tree
[c70733] by Gabor Melis Gabor Melis

0.9.8.13:
* don't flood the system with GCs in the :BINDING-STACK-GC-SAFETY
test as SLEEP totally loses track of time (really hurts on true
multiprocessor systems)
* log SLEEP bug

2006-01-05 20:00:39 Tree
[23c9f1] by Gabor Melis Gabor Melis

0.9.8.12:
* NetBSD has a different bug in SA_NODEFER, detect it (patch by
Richard M Kreuter)

2006-01-05 14:13:14 Tree
[3f1617] by Gabor Melis Gabor Melis

0.9.8.11:
* fix compiler looping on #p"" when compiled with high debug
settings (thanks to Xophe support)
* add the tests missed in .10

2006-01-05 14:04:23 Tree
[62b6c1] by Gabor Melis Gabor Melis

0.9.8.10:
* bug fix: allow non-simple string symbol names (reported by Paul
Dietz)

2006-01-05 11:17:58 Tree
[adeddf] by Nathan Froyd Nathan Froyd

0.9.8.9:
Merge "updated ppc callback patch and tests", Cyrus Harmon,
sbcl-devel 2006-01-01;
* ...with a fix from Heiner Schwarte, sbcl-devel 2005-12-25.

2006-01-04 14:52:29 Tree
[6c741e] by Rudi Schlatte Rudi Schlatte

0.9.8.8
Add reference to isatty() ; successful cross-build with this change

2006-01-03 14:45:54 Tree
[7fb597] by Christophe Rhodes Christophe Rhodes

0.9.8.7:
Merge "merge candidate 1" for SBCL/Win32.
... a lot done, a lot left to do.

2006-01-03 09:52:37 Tree
[cb2549] by Juho Snellman Juho Snellman

0.9.8.6:
Make the count of fasl header counted strings a 32-bit value
also on 64-bit platforms. This ensures that a x86-64 SBCL can
gracefully detect an error when trying to read a x86 fasl
instead of choking on a ridiculously long counted string.

2005-12-30 00:21:40 Tree
[8735f1] by Juho Snellman Juho Snellman

0.9.8.5:
Merge sbcl-devel "segmentation fault on recent FreeBSD-current"
by NIIMI Satoshi, 2005-28-12 (with minor changes).

Add comments to document the reasoning behind a couple of
odd constructs in the 0.9.8.3 EQUAL changes to make them
look less like CMUCL-style black magic.

2005-12-29 22:48:01 Tree
[89987c] by Christophe Rhodes Christophe Rhodes

0.9.8.4:
Beginnings of a Win32 merge.
... rearrange the build scripts to use input from files rather
than <<HERE documents.
... (no other changes; just working to get the meaty changes
isolated from the fluff)

2005-12-29 16:08:31 Tree
[3ceaa0] by Juho Snellman Juho Snellman

0.9.8.3:
Make EQUAL faster (about 50% improvement for short lists on x86-64).
As amazing as it might seem, there are actually real-world
applications where significant time is spent in EQUAL.

* Inline EQL in EQUAL
* Rearrange things a bit to enable the inlining
* Rewrite EQUAL to use a local helper function

2005-12-28 22:37:14 Tree
Older >