Emacs does not react on any command after error.

[kostafey]

When I evaluate wrong expresssion (wich case whatever error) after few seconds Emacs suddenly does not react on any command. Any activity is blocked.

Then I type C-g C-g C-g C-g C-g and I can work for a few seconds more before Emacs will stop again. Only nrepl restart (which should be executed quickly before emacs blocked again) solves the problem until next error occurs.

How can I debug such problem or provide more information?

• GNU Emacs 24.3.1
• CIDER 0.7.0alpha (package: 20140522.212)
• cider/cider-nrepl "0.7.0-SNAPSHOT" (the same problem with 0.6)

[bbatsov]

Seems like a problem with stacktrace middleware. You can monitor your *nrepl-events* buffer to understand what message leads to this unfortunate scenario. //cc @jeffvalk

[kostafey]

When I eval wrong expression SQLException occurs. After a few seconds Emacs stoped. The last expression ("id" "58" ... occurs only after C-g C-g C-g call.

Here is *nrepl-events* buffer tail:

("err" . "SQLException Wrong column type oracle.jdbc.driver.OraclePreparedStatement.setObjectCritical (OraclePreparedStatement.java:8921)\n")
 ("id" . "56")
 ("session" . "cb2c131a-a135-4d82-a1cd-0bac698ae5b1"))
(dict
 ("id" . "56")
 ("session" . "cb2c131a-a135-4d82-a1cd-0bac698ae5b1")
 ("status" "done"))
("id" "58" "op" "info" "session" "cb2c131a-a135-4d82-a1cd-0bac698ae5b1" "ns" "archive-server.sql-exec" "symbol" "get-blob")

[aamedina]

I've been running into this lately as well

[jeffvalk]

It sounds like this issue. Looking into this now. If anyone has a code snippet that will reproduce this, it would greatly help.

[jeffvalk]

When this happens, does the *nrepl-server cider-nrepl* buffer show a stacktrace?

[jeffvalk]

Or is there anything in *nrepl-connection cider-nrepl*?

[nathell]

I can reproduce this on current cider from melpa. My Emacs config is in https://github.com/nathell/dotemacsdotd.

I do M-x cider-jack-in in freshly started Emacs, then evaluate some expressions. Upon evaluating an erroneous expression (e.g., abc), cider doesn't print anything, not even the prompt. When I press RET, I get the prompt again.

The *nrepl-connection localhost* buffer is empty. *nrepl-events* doesn't show up when I do C-x b.

[bbatsov]

@nathell If you're Emacs doesn't freeze you don't have the same problem. Seems to me you haven't installed cider-nrepl. I should really add some warning in the REPL regarding this.

[kostafey]

@jeffvalk Not every error leads to such problem. Though I haven't found very easy way to reproduce it, here is the simplest app:

• Run lein new test-prj.
• Contents of project.clj

(defproject test-prj "0.1.0-SNAPSHOT"
  :description "FIXME: write description"
  :url "http://example.com/FIXME"
  :license {:name "Eclipse Public License"
            :url "http://www.eclipse.org/legal/epl-v10.html"}
  :dependencies [[org.clojure/clojure "1.5.1"]
                 [org.clojure/java.jdbc "0.3.3"]
                 [com.oracle/ojdbc6 "11.2.0"]])

• Contents of core.clj

(ns test-prj.core  
  (:require [clojure.java.jdbc :as j]))

(def oracle-test-db {:subprotocol "oracle"
                     :classname "oracle.jdbc.OracleDriver"
                     :subname "thin:@<host>:<port>/<db>"
                     :user "user"
                     :password "secret"})

(defn fail []
  (j/query oracle-test-db ["select field from table where id = ?" 1] 
           :result-set-fn
           (fn [rs]
             (print "RS: " rs))))

;; (fail)

field here is java.sql.Blob result-set object and id = 1 actually not exists. Eval (fail), wait for a few seconds and your Emacs will freeze.

If I'll find the easier way to reproduce it, I'll provide it immediately.

UPD: There is no record with such id. But I can't reproduce it with non-oracle database yet.

[kostafey]

• *nrepl-server test-prj* output nothing new:

nREPL server started on port 53241 on host 127.0.0.1

• *nrepl-connection test-prj* tail:
  (all lines from second line inclusive 2:id2:207:session36... occurs only after C-g C-g C-g call)

d2:ex27:class java.sql.SQLException2:id2:207:root-ex27:class java.sql.SQLException7:session36:7852adc1-ba25-4508-9cdf-ffd0b4c124be6:statusl10:eval-erroreed3:err158:SQLException Wrong column type  oracle.jdbc.driver.OraclePreparedStatement.setObjectCritical (OraclePreparedStatement.java:8921)
2:id2:207:session36:7852adc1-ba25-4508-9cdf-ffd0b4c124beed2:id2:207:session36:7852adc1-ba25-4508-9cdf-ffd0b4c124be6:statusl4:doneeed2:id2:217:session36:7852adc1-ba25-4508-9cdf-ffd0b4c124be5:valued5:added3:1.012:arglists-str10:([& sigs])6:columni1e3:doc168:params => positional-params* , or positional-params* & next-param
  positional-param => binding-form
  next-param => binding-form
  name => symbol

  Defines a function4:file100:jar:file:/home/kostafey/.m2/repository/org/clojure/clojure/1.5.1/clojure-1.5.1.jar!/clojure/core.clj9:forms-str63:  (fn name? [params*] exprs*)
  (fn name? ([params*] exprs*) +)4:linei4067e5:macro4:true4:name2:fn2:ns12:clojure.core8:resource16:clojure/core.clj12:special-form4:true3:url35:http://clojure.org/special_forms#fneed2:id2:217:session36:7852adc1-ba25-4508-9cdf-ffd0b4c124be6:statusl4:doneee

[kostafey]

I've found the easiest way to reproduce it! Just import something wrong.
e.g. literally:

(ns test-prj.core  
  (:import blah.blah))

[jeffvalk]

@kostafey Great find -- this is a big help! Digging into it now.

[jeffvalk]

Solved. See other discussion here.

And again, great repro code @kostafey; this was a huge help in tracking this down!

[bbatsov]

I just merged a fix for this (#577).