Hi JIT developers.
I've wrote 4 templates and about 20 tests to test Jitrino's internal
algorithms. I put them all into JIRA-1586
Also I've founded a problem with testing simplifier.cpp (test
testSimplifyIf). The problem was in
FlowGraph::foldBranch(ControlFlowGraph fg, Node* block,
On the 0x20F day of Apache Harmony Ivan Kollegov wrote:
Hi JIT developers.
I've wrote 4 templates and about 20 tests to test Jitrino's internal
algorithms. I put them all into JIRA-1586
Also I've founded a problem with testing simplifier.cpp (test
testSimplifyIf). The problem was in
Thank you Ivan.
The problem you found with TestRegistry.cpp will be fixed with the next my
patch for Jitrino testing framework.
I'll check all of your tests this weekend and post comments.
On 10/27/06, Ivan Kollegov [EMAIL PROTECTED] wrote:
Hi JIT developers.
I've wrote 4 templates and about
On 27 Oct 2006 18:43:46 +0700, Egor Pasko [EMAIL PROTECTED] wrote:
void ControlFlowGraph::removeNode(Nodes::iterator pos, bool erase) {
Node* node = *pos;
-assert(node!=entryNode);
+ if (node == entryNode) {
+ entryNode=NULL;
+ }
if (node == returnNode)
While you're at it, I'd like to get this patches in, but we probably
need a new target for building, as running normally with this framework
compiled in leads to problems.
Can we find a way to make it a different build target, or at least a flag
-Dwith.jitrino.test.framework
or something?
Geir,
I also thought about additional cmd-line param to compile and link
Jitrino.dll with tests. I hope we will have a lot of tests in future, so IMO
it's a good idea to separate it from default Jitrino build.
I'll try to provide this switch with the next patch and that could be a good
time to
On the 0x20F day of Apache Harmony Geir Magnusson, Jr. wrote:
While you're at it, I'd like to get this patches in, but we probably
need a new target for building, as running normally with this
framework compiled in leads to problems.
Can we find a way to make it a different build target, or
On 27 Oct 2006 20:15:27 +0700, Egor Pasko [EMAIL PROTECTED] wrote:
On the 0x20F day of Apache Harmony Geir Magnusson, Jr. wrote:
While you're at it, I'd like to get this patches in, but we probably
need a new target for building, as running normally with this
framework compiled in leads to
On the 0x20F day of Apache Harmony Mikhail Fursov wrote:
On 27 Oct 2006 20:15:27 +0700, Egor Pasko [EMAIL PROTECTED] wrote:
On the 0x20F day of Apache Harmony Geir Magnusson, Jr. wrote:
While you're at it, I'd like to get this patches in, but we probably
need a new target for building,
Egor Pasko wrote:
On the 0x20F day of Apache Harmony Geir Magnusson, Jr. wrote:
While you're at it, I'd like to get this patches in, but we probably
need a new target for building, as running normally with this
framework compiled in leads to problems.
Can we find a way to make it a different
yes, I saw - thanks for that.
I was trying to get 1531 to work yesterday, and I seemed to do something
wrong.
Can you look at 1531 and comment?
geir
Ivan Kollegov wrote:
Hi,
I opened new JIRA
http://issues.apache.org/jira/browse/HARMONY-1586
and attach my contribution
to keep things
Geir, my comment is in JIRA.
On 10/3/06, Geir Magnusson Jr. [EMAIL PROTECTED] wrote:
yes, I saw - thanks for that.
I was trying to get 1531 to work yesterday, and I seemed to do something
wrong.
Can you look at 1531 and comment?
geir
Ivan Kollegov wrote:
Hi,
I opened new JIRA
Ivan,
Thank you for the fix.
I checked the diff and it looks OK.
Actually, with this patch you remove the extra state from CFG - the special
behaviour of CFG without exit/entry nodes. And I think that this is right.
Now it's time to ask commiters to apply the patch.
On 10/3/06, Ivan Kollegov
so is my response :)
Mikhail Fursov wrote:
Geir, my comment is in JIRA.
On 10/3/06, Geir Magnusson Jr. [EMAIL PROTECTED] wrote:
yes, I saw - thanks for that.
I was trying to get 1531 to work yesterday, and I seemed to do something
wrong.
Can you look at 1531 and comment?
geir
Ivan
done
Mikhail Fursov wrote:
Ivan,
Thank you for the fix.
I checked the diff and it looks OK.
Actually, with this patch you remove the extra state from CFG - the special
behaviour of CFG without exit/entry nodes. And I think that this is right.
Now it's time to ask commiters to apply the patch.
Ivan,
I've checked all of the tests - this is very good beginning!
I see 2 directions to advance:
1) Integrate Jitrino tests into common harmony test system. So we can ask
people to use our test suite.
2) Start to test not only CFG but more complex IR examples. I mean to test
IR with instructions
Hello.
The documentation has helped me to write the tests.
I have written 4 templates of graphs and 20 tests for methods of
ControlFlowGraph.
I have founded a problem with test getNodeCountTest2. If there are no
nodes in graph I have follows
java.exe has encountered a problem and needs to
On Sep 26, 2006, at 9:25 AM, Ivan Kollegov wrote:
Hello.
The documentation has helped me to write the tests.
I have written 4 templates of graphs and 20 tests for methods of
ControlFlowGraph.
I have founded a problem with test getNodeCountTest2. If there are no
nodes in graph I have follows
Mikhail,
Great proposal! The IR-level JIT unit testing is not covered in
Harmony yet. I looked into the patch and it looks quite lightweight
though the test code would be difficult to understand without in-depth
JIT knowledge. Anyway, I think it would be possible to use such a
framework for unit
Hello JIT developers.
I am a student and want to learn JIT and understand its structure.
What should I learn first? Is there some documentation for JIT?
I think could write some tests for JITs methods. It would be useful for me.
Thanks,
Kollegov Ivan
On 9/22/06, Pavel Ozhdikhin [EMAIL
Hi Ivan!
Writing tests for internal JIT algorithms and optimizations is the best way
to learn Jitrino.OPT compiler from inside.
I think that we should start from testing the ControlFlowGraph related
algorithms first because of ControlFlowGraph structures and routines are
shared by both
Mikhail,
I like it :) This reminds me that we need to get the vmmagic regression
tests included as part of the JIT regression suite.
Weldon
On 9/21/06, Mikhail Fursov [EMAIL PROTECTED] wrote:
All,
This is the proposal to start the discussion and implementation of
Jitrino.OPT compiler
All,
This is the proposal to start the discussion and implementation of
Jitrino.OPT compiler internal testing framework.
There are a lot of optimizations in Jitrino with a variety of options but
only minor part of them is used and tested in the default configuration.
The result is when adding
23 matches
Mail list logo