The only bug I'm aware of right now is the stack problem in yyparse (which normally only shows when PIKE_DEBUG is enabled). That problem appears to be a bit tricky, though.
/ Martin Stjernholm, Roxen IS
Previous text:
2002-09-30 19:59: Subject: 7.3 -> 7.4
Are we ready to move the repository?
/ Peter Bortas
On Mon, Sep 30, 2002 at 11:00:07PM +0200, Martin Stjernholm, Roxen IS @ Pike developers forum wrote:
The only bug I'm aware of right now is the stack problem in yyparse (which normally only shows when PIKE_DEBUG is enabled). That problem appears to be a bit tricky, though.
has anyone looked at the crash caused by get_dir() with mor than 160000 files? (yes, i know, i never got around to properly reporting it in crunch.)
greetings, martin.
There's also a bug that Thread.pmod doesn't compile anymore.
/ Marcus Comstedt (ACROSS) (Hail Ilpalazzo!)
Previous text:
2002-09-30 22:58: Subject: 7.3 -> 7.4
The only bug I'm aware of right now is the stack problem in yyparse (which normally only shows when PIKE_DEBUG is enabled). That problem appears to be a bit tricky, though.
/ Martin Stjernholm, Roxen IS
Maybe you'd like to fix the testsuite too?
/pike/home/marcus/Pike/7.3/src/testsuite.in:3359: Test 761 (shift 1) (CRNL) failed. 0: mixed a() { 1: array data=({1, Thread.Mutex(), Thread.Condition(), 0}); 2: for(int e=0;e<3;e++) { data[3]++; thread_create(lambda(array data) { 3: for(int e=0;e<1000;e++) { object o=data[1]->lock(); data[0]*=2; 4: for(int d=0;d<5;d++) { data[0]--; data[0]*=2; } data[0]--; destruct(o); } 5: data[3]--; data[2]->signal();}, data);} 6: while(data[3]) data[2]->wait(); return data[0];; } 7: mixed b() { return 1; } 8:
Error: Bad argument 1 to condition->wait(). Expected object Unknown program: condition->wait() native module.Builtin.condition: wait() testsuite:7: a() /pike/home/marcus/Pike/7.3/bin/test_pike.pike:680: main(3,({"/pike/home/marcus/Pike/7.3/bin/test_pike.pike",0,0}))
/ Marcus Comstedt (ACROSS) (Hail Ilpalazzo!)
Previous text:
2002-10-01 10:19: Subject: 7.3 -> 7.4
Fixed.
/ Henrik Grubbström (Lysator)
Yes, I fixed it a few minutes ago.
/ Henrik Grubbström (Lysator)
Previous text:
2002-10-01 11:19: Subject: 7.3 -> 7.4
Maybe you'd like to fix the testsuite too?
/pike/home/marcus/Pike/7.3/src/testsuite.in:3359: Test 761 (shift 1) (CRNL) failed. 0: mixed a() { 1: array data=({1, Thread.Mutex(), Thread.Condition(), 0}); 2: for(int e=0;e<3;e++) { data[3]++; thread_create(lambda(array data) { 3: for(int e=0;e<1000;e++) { object o=data[1]->lock(); data[0]*=2; 4: for(int d=0;d<5;d++) { data[0]--; data[0]*=2; } data[0]--; destruct(o); } 5: data[3]--; data[2]->signal();}, data);} 6: while(data[3]) data[2]->wait(); return data[0];; } 7: mixed b() { return 1; } 8:
Error: Bad argument 1 to condition->wait(). Expected object Unknown program: condition->wait() native module.Builtin.condition: wait() testsuite:7: a() /pike/home/marcus/Pike/7.3/bin/test_pike.pike:680: main(3,({"/pike/home/marcus/Pike/7.3/bin/test_pike.pike",0,0}))
/ Marcus Comstedt (ACROSS) (Hail Ilpalazzo!)
After careful consideration and some discussion I think that those races doesn't need fixing. So just add a silly mutex lock to get something to feed Thread.Condition.wait. (It's not much of a performance issue at that point since the thread will wait anyway then.)
/ Martin Stjernholm, Roxen IS
Previous text:
2002-10-01 13:06: Subject: 7.3 -> 7.4
How about roxen.pike then?
/ Martin Nilsson (Fake Build Master)
Ok. There is one more wait in configuration.pike that you might want to take a look at as well.
/ Martin Nilsson (Fake Build Master)
Previous text:
2002-10-01 18:46: Subject: 7.3 -> 7.4
After careful consideration and some discussion I think that those races doesn't need fixing. So just add a silly mutex lock to get something to feed Thread.Condition.wait. (It's not much of a performance issue at that point since the thread will wait anyway then.)
/ Martin Stjernholm, Roxen IS
pike-devel@lists.lysator.liu.se