Closed catap closed 4 months ago
It is reproducible almost each switch ethernet -> wifi
To investigate that I've added that hack:
diff --git a/elmo/elmo-imap4.el b/elmo/elmo-imap4.el
index 74a1d825..bb84fd88 100644
--- a/elmo/elmo-imap4.el
+++ b/elmo/elmo-imap4.el
@@ -467,8 +467,13 @@ COMMAND is a list of command tokens or a single command token."
TAG is the tag of the command"
(with-current-buffer (process-buffer
(elmo-network-session-process-internal session))
+ (setq elmo-imap4-read-response-loop-counter 0)
(while (not (or (string= tag elmo-imap4-reached-tag)
(elmo-imap4-response-bye-p elmo-imap4-current-response)
+ (when
+ (> (setq elmo-imap4-read-response-loop-counter (1+ elmo-imap4-read-response-loop-counter)) 100)
+ (message "elmo-imap4-read-response iterates more than 100 times, abort")
+ t)
(when (elmo-imap4-response-garbage-p
elmo-imap4-current-response)
(message "Garbage response: %s"
And I have:
elmo-imap4-read-response iterates more than 100 times, abort
wl-biff: error ((IMAP error: No `OK' response from server.))
this time it requires to press C-g
, but at least I haven't lost emacs :)
But I need to re-connect via double pressing M-t
, otherwise emacs is seems blocked and C-g
is required to wake it up.
seems that this hack had broke wl-biff
Seems that C-g
is related to https://github.com/wanderlust/wanderlust/issues/148 and/or https://github.com/wanderlust/wanderlust/issues/213
I can't reproduce this issue on c15e8ece4f34f10479e17cda19d10b98f6be3ec1 anymore.
When I had running
wl
inside emacs I've disconnected from ethernet (and switched to wifi) which apparently changed machines IP and network.After that Emacs going to consume 100% CPU and connect to it via
lldb
allow to get this trace:Seems that
elmo-imap4-read-response
doesn't handle well EOF at some cases :(