summaryrefslogtreecommitdiff
path: root/hurd/debugging/trap_in_the_kernel.mdwn
diff options
context:
space:
mode:
Diffstat (limited to 'hurd/debugging/trap_in_the_kernel.mdwn')
-rw-r--r--hurd/debugging/trap_in_the_kernel.mdwn27
1 files changed, 27 insertions, 0 deletions
diff --git a/hurd/debugging/trap_in_the_kernel.mdwn b/hurd/debugging/trap_in_the_kernel.mdwn
new file mode 100644
index 00000000..11f989e3
--- /dev/null
+++ b/hurd/debugging/trap_in_the_kernel.mdwn
@@ -0,0 +1,27 @@
+[[!meta copyright="Copyright © 2010 Free Software Foundation, Inc."]]
+
+[[!meta license="""[[!toggle id="license" text="GFDL 1.2+"]][[!toggleable
+id="license" text="Permission is granted to copy, distribute and/or modify this
+document under the terms of the GNU Free Documentation License, Version 1.2 or
+any later version published by the Free Software Foundation; with no Invariant
+Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license
+is included in the section entitled [[GNU Free Documentation
+License|/fdl]]."]]"""]]
+
+[[!tag open_issue_gnumach open_issue_documentation]]
+
+IRC, #hurd, September 2010
+
+ <diegonc> when an application executes an out instruction in user mode, how is
+ kernel mode entered? general protection trap?
+ <youpi> some sort of trap, yes
+ <youpi> I'd rather think about illegal instruction, but yes
+ <diegonc> hm.. so to debug what happens inside that instruction I'll have to
+ break at the trap handler. Can I instruct kdb to stop only when a given task
+ caused the trap?
+ <youpi> applications usually don't trap, so what I usually do is to uncomment
+ the test at the end of user_trap() before the call to kdb_trap()
+ <diegonc> "if (debug_all_traps_with_kdb && .. " <- that test?
+ <youpi> yes
+ <youpi> so comment the test to make kdb_trap() called all the time
+ <diegonc> oh, I understand now :)