Search
Monday, 26th of September 2022, 1:07:58 UTC
2:32:59
Bike
random evening thought: compiling discriminators in a dedicated thread might be easier for making it work as well as for performance reasons
2:33:05
Bike
have to get it working to begin with first, tho
2:50:33
drmeister
We are blowing through the top of the stack.
2:53:55
Bike
hmm. the sjlj unwind ought to make sure the destination dynenv is on the stack before it does anything
3:02:44
drmeister
I mean - the stack pointer is exceeding the top of the stack.
3:14:19
Bike
like the bytecode stack pointer?
3:17:49
drmeister
https://usercontent.irccloud-cdn.com/file/7SAfcNZH/image.png
3:35:50
drmeister
nargs words (arguments) are pushed onto the stack and then bytecode_call returns with those words popped off the stack - correct?
3:39:45
drmeister
Uh return values.
3:43:27
drmeister
Bike: Is there a simple relationship between the vm._stackPointer on entry to bytecode_call and on normal return?
9:12:27
Guest1362
** NICK 040AAGS3Q
11:18:28
attila_lendvai_
** NICK attila_lendvai
11:38:45
drmeister
There is a lot of stuff being left on the stack...
11:38:48
drmeister
https://www.irccloud.com/pastebin/f8bDL55m/
11:39:31
drmeister
This is when ASDF is performing `ASDF/SOURCE-REGISTRY:INITIALIZE-SOURCE-REGISTRY`
11:39:58
drmeister
It works by trying to open files and if they don't exist it signals an error and the stack is supposed to unwind.
11:40:12
drmeister
But the vm._stackPointer isn't being updated.
11:41:08
drmeister
I added a guard to the vm stack so if we attempt to write into it then it segfaults.
11:41:42
drmeister
This is to the end of the stack - now it won't overwrite other structures and give confusing errors.
11:42:22
drmeister
The dump above is a dump of consecutive words on the stack
11:42:46
drmeister
::notify Bike Can you check the logs this morning.
11:42:46
Colleen
drmeister: Got it. I'll let Bike know as soon as possible.
12:33:43
Colleen
Bike: drmeister said 50 minutes, 57 seconds ago: Can you check the logs this morning.
12:35:21
Bike
the stack pointer is not being updated when we do a nonlocal exit?
12:42:06
drmeister
In the vm-boehm branch
12:42:45
drmeister
I would have found it earlier had I had the guards in place.
12:42:51
Bike
does the nonlocal exit originate from compiled or bytecoded code?
12:44:44
Bike
it's going to be pretty annoying if we have to put the unwind-protect in bytecode_call back.
12:47:56
Bike
huh. well that's where i'd expect the stack pointer to get updated properly. although... maybe unwind protect won't do it so well...
12:48:25
drmeister
This is OPEN, signaling an error because the file doesn't exist.
12:49:33
drmeister
https://www.irccloud.com/pastebin/V1bXs4gq/
12:49:51
drmeister
It's OPEN - with a file that doesn't exist.
12:50:05
drmeister
That invokes error, that invokes UNIVERSAL-ERROR-HANDLER
12:50:14
drmeister
That invokes %SIGNAL (compiled)
12:50:48
drmeister
that does a funcall into a bytecode wrapper and that invokes a compiled LAMBDA - line 2.
12:50:56
drmeister
That does the cc_sjlj_unwind
12:51:35
drmeister
So that means compiled - right?
12:54:48
Bike
and then it goes into unwind-protect cleanup(s), right?
13:06:42
kaskal
Hi! I'm new in this channel, I'd like to use clasp to interface it with our quantum chemistry code in c++, but I have some doubts. So we normally use the intel compiler for production calculations, is this possible to use clasp with code compiled with icc? How easy is it to compile clasp in HPC systems with the limited amount of libraries out of the box ?
Monday, 26th of September 2022, 13:07:58 UTC