Qian Yun
unread,6:21 AM (11 hours ago) 6:21 AMSign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to fricas-devel
I hit this strange situation when I was testing
recycle-kernel-cache patch with a long running session:
to compute 10000 integrals in a row.
I use TimeoutPackage package with uses sb-ext:with-timeout,
to avoid stuck at an integral for too long.
The strange situation is, after over 9000 integrals computed,
the session is filled with errors complaining
"WEAK POINTER is not of type BOOT::SPAD_KERNEL".
I guess that in the middle of recycling kernel cache,
timeout is reached, and kernel cache is left at inconsistent
state. (the array contains weak pointer instead of kernel.)
The SBCL documentation does say that sb-ext:with-timeout
can be very dangerous.
This is also a remainder that sb-ext:with-timeout can cause
inconsistency elsewhere in the code base silently.
I'll try if WITHOUT-INTERRUPTS can help in this case.
- Qian