Nicholas Clark [EMAIL PROTECTED] wrote:
How come most tinderboxes kept going without failing? What's making the
choice on the value of ARENA_DOD_FLAGS ?
ARENA_DOD_FLAGS is turned on by default. If there is no memalign or such
library function (which it depends on), this define is disabled.
On Sat, Jan 10, 2004 at 04:25:58PM -0800, Jeff Clites wrote:
allocate chunks of memory with arbitrary power-of-2 alignment. So all
the platforms being tested on the tinders probably have this. (Of
course, you can manually set ARENA_DOD_FLAGS to false in the source,
for testing.)
My
Luke Palmer [EMAIL PROTECTED] wrote:
Leopold Toetsch writes:
It seems, that in that case we still need to walk the PMC arenas and
reset all live bits. OTOH when ARENA_DOD_FLAGS is on, this isn't too
expensive because it can be done my masking a full word worth of 8 PMCs
at once. So its still
On Sat, Jan 10, 2004 at 02:28:38PM +0100, Leopold Toetsch wrote:
Luke Palmer [EMAIL PROTECTED] wrote:
Leopold Toetsch writes:
It seems, that in that case we still need to walk the PMC arenas and
reset all live bits. OTOH when ARENA_DOD_FLAGS is on, this isn't too
expensive because it can
Nicholas Clark [EMAIL PROTECTED] wrote:
src/dod.c: In function `clear_live_bits':
src/dod.c:755: `cur_arena' undeclared (first use in this function)
The appended patch cures it (and all tests pass) but I'm not sure if it
is correct.
Ah yep, thanks. I didn't test the patch with
On Sat, Jan 10, 2004 at 08:39:51PM +0100, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
src/dod.c: In function `clear_live_bits':
src/dod.c:755: `cur_arena' undeclared (first use in this function)
The appended patch cures it (and all tests pass) but I'm not sure if it
On Jan 10, 2004, at 11:54 AM, Nicholas Clark wrote:
On Sat, Jan 10, 2004 at 08:39:51PM +0100, Leopold Toetsch wrote:
Nicholas Clark [EMAIL PROTECTED] wrote:
src/dod.c: In function `clear_live_bits':
src/dod.c:755: `cur_arena' undeclared (first use in this function)
The appended patch cures it
Jeff Clites [EMAIL PROTECTED] writes:
It gets set in include/parrot/pobj.h, and is basically set to true if
your platform has some flavor of memalign(), which allows you to
allocate chunks of memory with arbitrary power-of-2 alignment. So all
the platforms being tested on the tinders probably
Leopold Toetsch writes:
Luke Palmer [EMAIL PROTECTED] wrote:
... I'm not
against somebody else maintaining the patch in the meantime :-)
I went again through the patch and the original one from Sept, 5th. But
it seems that one thing is missing in both:
*If* all PMCs which
Luke Palmer [EMAIL PROTECTED] wrote:
Leopold Toetsch writes:
Moving the needs_early_DOD_FLAG out of the arena_flags is suboptimal
(and probably the reason for the 5% slowdown for the eager case). Now the
relevant flags is the high_priority_DOD_FLAG. If I get the patch right,
it gets set on
Luke Palmer [EMAIL PROTECTED] wrote:
... I'm not
against somebody else maintaining the patch in the meantime :-)
I went again through the patch and the original one from Sept, 5th. But
it seems that one thing is missing in both:
*If* all PMCs which needs_early_DOD are seen live, the DOD run
Leopold Toetsch writes:
Luke Palmer [EMAIL PROTECTED] wrote:
After many months of lying dormant, I figured I'd get my act together
and adapt this patch to the few recent modifications. And this time,
I'm posting a benchmark!
Wow, thanks.
Some comments:
-
After many months of lying dormant, I figured I'd get my act together
and adapt this patch to the few recent modifications. And this time,
I'm posting a benchmark!
My results get about 5% slowdown in the eager case, and the usual
10,000% speedup in the lazy case.
Luke
First, the benchmark
Luke Palmer [EMAIL PROTECTED] wrote:
After many months of lying dormant, I figured I'd get my act together
and adapt this patch to the few recent modifications. And this time,
I'm posting a benchmark!
Wow, thanks.
Some comments:
-b_PObj_needs_early_DOD_FLAG = 1 27,
+/* true if
Luke Palmer [EMAIL PROTECTED] wrote:
One more remark:
+if (PObj_needs_early_DOD_TEST(obj))
+++interpreter-num_early_PMCs_seen;
When is this counter reset?
leo
On Jan 5, 2004, at 5:47 AM, Luke Palmer wrote:
After many months of lying dormant, I figured I'd get my act together
and adapt this patch to the few recent modifications. And this time,
I'm posting a benchmark!
My results get about 5% slowdown in the eager case, and the usual
10,000% speedup in
Jeff Clites writes:
On Jan 5, 2004, at 5:47 AM, Luke Palmer wrote:
After many months of lying dormant, I figured I'd get my act together
and adapt this patch to the few recent modifications. And this time,
I'm posting a benchmark!
My results get about 5% slowdown in the eager case, and
17 matches
Mail list logo