debugger will not step in to a function after stepping out of another function
Categories
(DevTools :: Debugger, defect, P2)
Tracking
(firefox70 fixed)
Tracking | Status | |
---|---|---|
firefox70 | --- | fixed |
People
(Reporter: bugzilla, Assigned: jlast)
References
(Blocks 1 open bug)
Details
(Whiteboard: [debugger-mvp])
Attachments
(2 files, 1 obsolete file)
Reporter | ||
Updated•10 years ago
|
Comment 1•9 years ago
|
||
Comment 2•9 years ago
|
||
Comment 3•9 years ago
|
||
Comment 4•9 years ago
|
||
Updated•6 years ago
|
Assignee | ||
Comment 5•6 years ago
|
||
STR
- go to https://appear.in/firefox-debugger
- step-in
ER: inside of C
AR: inside of A
I think this is because when a call is at the beginning of a statement we place the breakpoint at the statement..
NOTE: chrome places the bp at the call.
NOTE: I can be persuaded that what we're doing is right because we stop at the beginning, but now that we have column bps i feel like we can be more opinionated. e.g. if the user wants to stop at a() they can.
Assignee | ||
Updated•6 years ago
|
Assignee | ||
Comment 7•6 years ago
|
||
ha looks like the case :)
https://dbg-step-in.glitch.me/
Assignee | ||
Updated•6 years ago
|
Assignee | ||
Updated•6 years ago
|
Comment 8•6 years ago
|
||
Yeah this seems more like a question of what we want to do than anything. This is expected given the rules we laid out for always treating the first opcode in an statement-level expression as the place to stop, but if we want to change that to special-case it if the first position is a call and used the call opcode instead that seems fine to me.
All of the SpiderMonkey position stuff is gone from my head again so I'd have to dig in to see how complicated it would be, but it might not be too bad?
Assignee | ||
Comment 9•6 years ago
|
||
I think using the call opcode makes sense.
Comment 10•6 years ago
|
||
One alternative we could consider would be to put the call marker on the (
when a function call is specifically at the start like this so when you do
foo(bar());
// ^ ^^
so the first ^
would be the expression start, the second would be the call to foo
and the third would be the call to bar
. That way we still keep the current behavior of always having the expressions tart be a breakable location. Personally I do like that we currently do that. That said, I don't feel super strongly necessarily.
Assignee | ||
Comment 11•6 years ago
|
||
that seems reasonable, i don't think people will notice that call has that special case. I'm ambivalent between these two options.
Updated•6 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 12•5 years ago
|
||
Logan, do you think this might be a nice first C++ bug for me?
Assignee | ||
Comment 13•5 years ago
|
||
here are some notes: https://gist.github.com/569ec70040abc74b5f85471a06370bbd
and a link to my failing build: https://gist.github.com/jasonLaster/049323fc81122e65a2837eb244e6431e
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 14•5 years ago
|
||
Updated•5 years ago
|
Assignee | ||
Comment 15•5 years ago
|
||
Comment 16•5 years ago
|
||
Comment 17•5 years ago
|
||
bugherder |
Assignee | ||
Comment 18•5 years ago
|
||
Updated•5 years ago
|
Comment 19•5 years ago
|
||
Comment 20•5 years ago
|
||
Backed out for failures on browser_dbg_rr_breakpoints-03.js
Backout link: https://hg.mozilla.org/integration/autoland/rev/4759a3899a17cfc46576aef3fec4f6f88bc2571d
Log link: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=260506575&repo=autoland&lineNumber=1019
Comment 21•5 years ago
|
||
Comment 22•5 years ago
|
||
bugherder |
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Description
•