on compile slime causes kernel to run out of memory
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| slime (Ubuntu) |
Invalid
|
Undecided
|
Unassigned | ||
Bug Description
Binary package hint: slime
steps to reproduce:
I was able to reproduce this using both gnu-emacs and xemacs:
load up slime using M-x slime.
then, do this at the slime prompt
(asdf:oos 'asdf:load-op :aserve)<enter>
then:
(asdf:oos 'asdf:load-op :parenscript)
no prompt is returned.
This doesn't happen when I run these commands using cmucl in a terminal (Konsole) window
I managed to retrieve this from the logs. Not sure if it's related, but it may be
Jun 16 01:54:34 jordan-laptop kernel: [58850.476000] Out of memory: kill process
13046 (lisp) score 328815 or a child
Jun 16 01:54:34 jordan-laptop kernel: [58850.476000] Killed process 13046 (lisp)
also, upon exiting either of the emacsen, the lisp process is not killed and must be killed manually.
This may be two seperate bugs, but I am not sure
| Jordan Kanter (jikanter) wrote addendum: slime memory leak | #1 |
| Jordan Kanter (jikanter) wrote | #2 |
| Jeff Anderson (jander99) wrote | #3 |
| Changed in slime (Ubuntu): | |
| status: | New → Incomplete |
| Jordan Kanter (jikanter) wrote Re: [Bug 121881] Re: on compile slime causes kernel to run out of memory | #4 |
| xteejx (xteejx-deactivatedaccount) wrote | #5 |
| Changed in slime (Ubuntu): | |
| status: | Incomplete → Invalid |
Ok, so after running it through gdb, and forcing the function to return with 'return',
slime returns a segfault to emacs through the slime buffer.
I have attached the gdb session as well as a screenshot of the emacs session.
note that I triggered this particular hang by hitting the <tab> key, so it is hard to determine exactly what slime command is causing these hangs.
I have also attached an strace of the process to it. it seems like the system process hangs up at around line 2219 where the poll() system call is interrupted with an EINTR.
Not sure how helpful that is.