227 Commits

Author SHA1 Message Date
Vladimir Kozlov
63c98ed888 6746892: Register Allocator does not process a data phi with one unique input correctly
Always look for the existing phi for a processed live_range.

Reviewed-by: rasbold
2008-09-10 14:29:32 -07:00
Vladimir Kozlov
8174b8b497 6676462: JVM sometimes would suddenly consume significant amount of memory
Add asserts with dead loop checks in AddNode::Ideal().

Reviewed-by: never
2008-09-05 13:33:55 -07:00
Tom Rodriguez
3d9b153464 6738933: assert with base pointers must match with compressed oops enabled
Reviewed-by: kvn, rasbold
2008-09-02 15:03:05 -07:00
Chuck Rasbold
e4e8777d6b Merge 2008-09-02 06:55:23 -07:00
Tom Rodriguez
aba715ae04 6732698: crash with dead code from compressed oops in gcm
Reviewed-by: rasbold
2008-08-28 23:03:55 -07:00
Chuck Rasbold
bdd62705f2 6611837: block frequency is zero
Insert_goto_at should set frequency for newly created blocks

Reviewed-by: never
2008-08-28 10:22:12 -07:00
Vladimir Kozlov
8ab0a4d7a9 6732732: CTW with EA: assert(n != 0L,"Bad immediate dominator info.")
Missing edge to a call's return value in EA Connection Graph.

Reviewed-by: never
2008-08-27 16:33:34 -07:00
Vladimir Kozlov
3c4879eb9a 6741738: TypePtr::add_offset() set incorrect offset when the add overflows
Set offset to OffsetBot when the add overflows in TypePtr::add_offset()

Reviewed-by: jrose, never
2008-08-27 14:47:32 -07:00
Vladimir Kozlov
37306315b2 6736417: Fastdebug C2 crashes in StoreBNode::Ideal
The result of step_through_mergemem() and remove_dead_region() is not checked in some cases.

Reviewed-by: never
2008-08-27 09:15:46 -07:00
Tom Rodriguez
26c780da72 5108146: Merge i486 and amd64 cpu directories
6459804: Want client (c1) compiler for x86_64 (amd64) for faster start-up

Reviewed-by: kvn
2008-08-27 00:21:55 -07:00
Matthias Klose
2697216f3a 6741642: bad enum definition in ciTypeFlow.hpp
Reviewed-by: rasbold, martin
2008-08-26 15:49:40 -07:00
Y. Srinivas Ramakrishna
5e6c038292 Merge 2008-08-25 12:16:34 -07:00
Chuck Rasbold
fe1d1630dc Merge 2008-08-21 05:49:06 -07:00
Jon Masamitsu
befe056b42 6728478: Assertion at parallel promotion from young to old generation
The fix avoids a call to address_for_index() in this particular situation where it is not known if the passed index is in bounds.

Reviewed-by: tonyp
2008-08-20 23:05:04 -07:00
Y. Srinivas Ramakrishna
6804a4e8ef 6739357: CMS: Switch off CMSPrecleanRefLists1 until 6722113 can be fixed
Temporarily switch off the precleaning of Reference lists completely until related issues are fixed in 6722113.

Reviewed-by: jmasa, poonam, tonyp
2008-08-20 15:41:36 -07:00
Andrei Pangin
f0d12afc79 Merge 2008-08-20 12:24:29 -07:00
Y. Srinivas Ramakrishna
b298e3e210 Merge 2008-08-20 11:23:58 -07:00
Y. Srinivas Ramakrishna
1d45ee2056 Merge 2008-08-20 01:30:32 -07:00
Antonios Printezis
b00baea4a1 6736341: PermGen size is insufficient for jconsole
Removing two buggy methods that should not be used, but ended up being used due to a re-organization in the class hierarchy.

Reviewed-by: jmasa, ysr, kamg, coleenp
2008-08-19 17:55:15 -04:00
Chuck Rasbold
ae4e755124 6730716: nulls from two unrelated classes compare not equal
Check for not-nullness after proving that types are unrelated

Reviewed-by: kvn, never
2008-08-19 07:25:02 -07:00
Andrei Pangin
059e448264 Merge 2008-08-19 06:02:07 -07:00
Tom Rodriguez
45f8e241e3 6732194: Data corruption dependent on -server/-client/-Xbatch
Rematerializing nodes results in incorrect inputs

Reviewed-by: rasbold
2008-08-18 23:17:51 -07:00
Xiaobin Lu
cf5e518aff 6459085: naked pointer subtractions in class data sharing code
Reviewed-by: jcoomes
2008-08-18 14:53:36 -07:00
Xiaobin Lu
bf2763ee4e 6608862: segv in JvmtiEnvBase::check_for_periodic_clean_up()
Reviewed-by: dholmes, dcubed, jcoomes
2008-08-15 10:08:20 -07:00
Y. Srinivas Ramakrishna
57dabb4fcb 6722112: CMS: Incorrect encoding of overflown object arrays during concurrent precleaning
When an object array overflows during precleaning, we should have been marking the entire array dirty, not just its first card.

Reviewed-by: jmasa, poonam, tonyp
2008-08-14 17:58:35 -07:00
Tomas Hurka
27a4da4686 6625846: Export system property java.version via jvmstat
Java.version added to property_counters_ss array

Reviewed-by: swamyv
2008-08-14 21:05:51 +02:00
Chuck Rasbold
1e49d39e6b 6732154: REG: Printing an Image using image/gif doc flavor crashes the VM, Solsparc
Delay transform call until uses of t2 are constructed

Reviewed-by: never
2008-08-14 10:15:29 -07:00
Erik Trimble
8242ae771a Merge 2008-08-10 21:31:42 -07:00
Jon Masamitsu
b28ffeff6e Merge 2008-08-04 12:15:55 -07:00
Jon Masamitsu
b57ca9ed91 Merge 2008-08-04 12:11:36 -07:00
Vladimir Kozlov
897278fce0 6731641: assert(m->adr_type() == mach->adr_type(),"matcher should not change adr type")
Fixed few addP node type and narrow oop type problems.

Reviewed-by: rasbold, never
2008-08-01 10:06:45 -07:00
Vladimir Kozlov
5e94e3bf69 6732312: Switch off executing Escape Analysis by default
Switch off executing Escape Analysis by default for now

Reviewed-by: rasbold
2008-07-31 15:47:42 -07:00
Vladimir Kozlov
73379adb82 Merge 2008-07-31 13:42:21 -07:00
Jon Masamitsu
74b3de5c3b 6730514: assertion failure in mangling code when expanding by 0 bytes
An expansion by 0 bytes was not anticipated when the assertion was composed.

Reviewed-by: jjh, jcoomes, apetrusenko
2008-07-30 11:54:00 -07:00
Vladimir Kozlov
6ff4206fe4 Merge 2008-07-30 09:54:21 -07:00
Chuck Rasbold
d35e1bbd88 6730192: expression stack wrong at deoptimization point
Add safepoint before popping expression stack, not after

Reviewed-by: kvn
2008-07-29 14:48:25 -07:00
Vladimir Kozlov
50c4a23cac 6726999: nsk/stress/jck12a/jck12a010 assert(n != null,"Bad immediate dominator info.")
Escape Analysis fixes.

Reviewed-by: never, rasbold
2008-07-28 17:12:52 -07:00
Jon Masamitsu
b8633a3b01 Merge 2008-07-28 15:30:23 -07:00
Keith McGuigan
74faacc945 6721093: -XX:AppendRatio=N not supported
Add mechanism to ignore unsupported flags for a set period of time

Reviewed-by: acorn, never, coleenp
2008-07-28 14:07:44 -04:00
Tom Rodriguez
20de63f383 6700047: C2 failed in idom_no_update
Partial peeling shouldn't place clones into loop

Reviewed-by: kvn
2008-07-25 15:54:23 -07:00
Tom Rodriguez
7b4a7f7bdc 6712835: Server compiler fails with assertion (loop_count < K,"infinite loop in PhaseIterGVN::transform")
Reviewed-by: kvn
2008-07-25 11:32:56 -07:00
Erik Trimble
632c83fb46 6732819: Turn off compressed oops by default for now
Workaround for CompOops bug

Reviewed-by: coleenp
2008-08-01 18:51:27 -07:00
Chuck Rasbold
fcbf2d1245 6726504: handle do_ifxxx calls in parser more uniformly
Make do_ifnull() handling similar to do_if()

Reviewed-by: jrose, kvn
2008-07-21 13:37:05 -07:00
Coleen Phillimore
20dba03e99 6716785: implicit null checks not triggering with CompressedOops
Allocate alignment-sized page(s) below java heap so that memory accesses at heap_base+1page give signal and cause an implicit null check

Reviewed-by: kvn, jmasa, phh, jcoomes
2008-07-19 17:38:22 -04:00
Igor Veresov
1193f0f9db 6723228: NUMA allocator: assert(lgrp_id != -1, "No lgrp_id set")
6723229: NUMA allocator: assert(lgrp_num > 0, "There should be at least one locality group")

The fix takes care of the assertion triggered during TLAB resizing after reconfiguration. Also it now handles a defect in the topology graph, in which a single leaf node doesn't have memory.

Reviewed-by: jmasa
2008-07-17 10:26:33 -07:00
Vladimir Kozlov
fae39068e8 6723160: Nightly failure: Error: meet not symmetric
Add missing _instance_id settings and other EA fixes.

Reviewed-by: rasbold
2008-07-16 16:04:39 -07:00
Chuck Rasbold
a8fc1db8c1 6707044: uncommon_trap of ifnull bytecode leaves garbage on expression stack
Remove call to repush_if_args()

Reviewed-by: kvn, jrose
2008-07-16 10:08:57 -07:00
John Coomes
26625ab7f5 6724367: par compact could clear less young gen summary data
Reviewed-by: jmasa, apetrusenko
2008-07-11 16:11:50 -07:00
John Coomes
038a30dde3 6483129: par compact assertion failure (new_top > bottom)
Avoid computing the dense prefix if a space is empty

Reviewed-by: pbk, tonyp
2008-07-11 16:11:43 -07:00
John Coomes
398aa66a77 6718283: existing uses of *_FORMAT_W() were broken by 6521491
Reviewed-by: ysr, pbk
2008-07-11 16:11:34 -07:00