Skip to contentSkip to conversation

disable "Call Trace:"

DISCUSSION 積丹尼 Dan Jacobson 2 participants 4 messages Aug 8, 2014 – Aug 10, 2014

lore
Share & tools
lore
#0 lore raw

Idea: if the user could disable
the "Call Trace:"
seen upon "Kernel panic - not syncing: attempted to kill init!" he would then be able to see more lines above without having it shoved off the screen with the useless (to him) call trace.

e.g., using panic=222 call_trace=disabled would give him enough time and room to read more of what happened from the screen.

#1 lore raw
Quoted 5 lines · 積丹尼 Dan Jacobson

Idea: if the user could disable
the "Call Trace:"
seen upon "Kernel panic - not syncing: attempted to kill init!" he would then be able to see more lines above without having it shoved off the screen with the useless (to him) call trace.

When would you want to disable it? at kernel boot time (as a boot option) or sometime later?

Quoted 3 lines

e.g., using panic=222 call_trace=disabled would give him enough time and room to read more of what happened from the screen.
--

Makes some sense to me and it's not difficult to implement.

But would you only want to disable Call Trace: on panics or any time that a call trace might be printed? and would disabling call trace also disable the raw stack dump and Code: output, or literally just the call trace?

Signature · 1 line
~Randy
#2 lore raw
In reply to 積丹尼 Dan Jacobson
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"RD" == Randy Dunlap <[email protected]> writes:

RD> When would you want to disable it? at kernel boot time (as a boot option) RD> or sometime later?

RD> But would you only want to disable Call Trace: on panics or any time that a RD> call trace might be printed? and would disabling call trace also disable RD> the raw stack dump and Code: output, or literally just the call trace?

All I know is e.g., somehow something went wrong and the user finds himself unable to boot and unable to read the reason, as Linux is famous for not having a way to scroll messages that have flown off the screen and boot_delay= doesn't work good and he doesn't own another terminal he can plug into a port or even have a port etc. etc. etc.

And he is sitting there and there is the reason, the main body of which has just flown off the screen.

All pushed off due to this (useless to him) "Call Trace:" and some other hex stuff lying above "Call Trace".

So if only he could please be able to, there hitting "e" in Grub, be able to add so "call_trace=no_thanks" "other_hex_stuff=no_thanks" to that linux command line, so then on the next run there will be something more left upon the screen! Thank you!

#3 lore raw
In reply to 積丹尼 Dan Jacobson

Here's another one I saw occupying the screen:

[...] ---[ end trace ... ]---
[...] Fixing recursive fault but reboot is needed!

Make sure that the user can also turn that one off and run it again to reveal the messages that were pushed off the screen above it.