bug#27036: 25.1; Blocking call to accept-process-output with quit inhibited!!

[Dmitry Gutov]

On 5/24/17 10:05 PM, Eli Zaretskii wrote:

> If you can reproduce this at will, run under GDB, put a breakpoint at
> the line which emits the message, and when it breaks, produce a
> backtrace.  I'd expect the offending code to be evident from that,
> especially if you also run "xbacktrace" to see the Lisp backtrace.

Here it is:

Thread 1 "emacs" hit Breakpoint 3, wait_reading_process_output 
(time_limit=0, nsecs=0, read_kbd=read_kbd@entry=0, 
do_display=do_display@entry=false, 
wait_for_cell=wait_for_cell@entry=XIL(0),
    wait_proc=wait_proc@entry=0x1399c30 <bss_sbrk_buffer+7942096>, 
just_wait_proc=0) at process.c:5015
5015	    message1 ("Blocking call to accept-process-output with quit 
inhibited!!");
(gdb) xbacktrace
"accept-process-output" (0xffffb980)
"inferior-octave-send-list-and-digest" (0xffffbc78)
"octave-eldoc-function-signatures" (0xffffbf68)
"octave-eldoc-function" (0xffffc410)
"apply" (0xffffc408)
0x13fef90 PVEC_COMPILED
"eldoc-print-current-symbol-info" (0xffffc950)
0xa3dc38 PVEC_COMPILED
"apply" (0xffffcd58)
"timer-event-handler" (0xffffd0a8)

Any ideas?

eldoc-print-current-symbol-info uses with-demoted-errors, which expands 
into a condition-case, but does that inhibit quitting somehow?

And inferior-octave-send-list-and-digest uses unwind-protect.