Traceback tables should remain on for Mac OS X release builds

RESOLVED FIXED in mozilla0.9.6

Status

SeaMonkey
Build Config
P2
normal
RESOLVED FIXED
16 years ago
13 years ago

People

(Reporter: Simon Fraser, Assigned: J.J. Enser)

Tracking

Trunk
mozilla0.9.6
PowerPC
Mac OS X
Dependency tree / graph

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: 10/10 needs r=, sr=, [PDT+] and [OSX+])

Attachments

(6 attachments)

(Reporter)

Description

16 years ago
We should leave traceback tables on for Mac OS X release builds, otherwise we 
have no way to get info about crashes (since we have no Talkback).

Comment 1

16 years ago
nsbranch+, we want this in our next release.
Keywords: nsbranch+
Priority: -- → P2
Whiteboard: 9/28 need fix and eventual PDT+ and OSX+
Target Milestone: --- → mozilla0.9.5

Updated

16 years ago
Blocks: 102989
(Assignee)

Comment 2

16 years ago
The same IDE_Options.h file is used for Classic & Carbon builds, so if we leave
traceback option turned on, it will affects both builds. is that ok ?
Status: NEW → ASSIGNED
(Reporter)

Comment 3

16 years ago
How about #if TARGET_CARBON?

Comment 4

16 years ago
not blocking 0.9.5, moving to 0.9.6

Target Milestone: mozilla0.9.5 → mozilla0.9.6

Updated

16 years ago
Blocks: 102998
(Assignee)

Comment 5

16 years ago
Created attachment 52939 [details] [diff] [review]
patch to IDE_Options.h to leave tracback option on for MacOSX build (#ifdef TARGET_CARBON)

Updated

16 years ago
Whiteboard: 9/28 need fix and eventual PDT+ and OSX+ → 9/28 need fix and eventual PDT+, [OSX+]

Comment 6

16 years ago
jj, you need to make that a unified diff with a few lines of context.  Use the 
Advanced Diff option in MacCVS Pro
(Assignee)

Comment 7

16 years ago
Fine, I'll post a context diff even though this one is pretty straightforward :-/

seeking review for the patch to IDE_Options.
This needs to land on the 0.9.4 branch first, then everywhere else.

note: I have an equivalent patch for ns/build/mac/IDE_Options.h. I suppose both
are needed
Whiteboard: 9/28 need fix and eventual PDT+, [OSX+] → 10/10 needs r=, sr=, PDT+ and OSX+
(Assignee)

Comment 8

16 years ago
Created attachment 52945 [details] [diff] [review]
Same patch as above, but as "unified diff"
(Reporter)

Comment 9

16 years ago
Please left-align the #ifdef TARGET_CARBON. And it should be
#if TARGET_CARBON

Comment 10

16 years ago
Created attachment 52955 [details] [diff] [review]
Additional patch required to make the #if TARGET_CARBON test work in IDE_Options.h

Comment 11

16 years ago
Comment on attachment 52955 [details] [diff] [review]
Additional patch required to make the #if TARGET_CARBON test work in IDE_Options.h

r=sdagley w/smfr's changes.  You'll also need the patch I attached or the #if TARGET_CARBON test will always fail in IDE_Options.h
Attachment #52955 - Flags: review+

Comment 12

16 years ago
Comment on attachment 52945 [details] [diff] [review]
Same patch as above, but as "unified diff"

This is the one I meant to r=
Attachment #52945 - Flags: review+
(Reporter)

Comment 13

16 years ago
Comment on attachment 52955 [details] [diff] [review]
Additional patch required to make the #if TARGET_CARBON test work in IDE_Options.h

This is wrong. IDE_Options.h has to come after DefinesMac.h.
Attachment #52955 - Flags: needs-work+

Comment 14

16 years ago
um, that's what patch 52955 does

Comment 15

16 years ago
Never mind, I can't read
(Reporter)

Comment 16

16 years ago
Comment on attachment 52955 [details] [diff] [review]
Additional patch required to make the #if TARGET_CARBON test work in IDE_Options.h

I suck, this is OK.
Attachment #52955 - Flags: needs-work+ → superreview+

Updated

16 years ago
Whiteboard: 10/10 needs r=, sr=, PDT+ and OSX+ → 10/10 needs r=, sr=, [PDT+] and [OSX+]
(Assignee)

Comment 17

16 years ago
Created attachment 53030 [details] [diff] [review]
rev#3 of the patch to IDE_Options.h with Simon's comments
(Assignee)

Comment 18

16 years ago
Simon, Steve, please add your respective r/sr marks for latest patch (attachment
53030 [review])

Comment 19

16 years ago
Comment on attachment 53030 [details] [diff] [review]
rev#3 of the patch to IDE_Options.h with Simon's comments

r=sdagley
Attachment #53030 - Flags: review+
(Reporter)

Comment 20

16 years ago
Comment on attachment 53030 [details] [diff] [review]
rev#3 of the patch to IDE_Options.h with Simon's comments

Fix the spacing, and sr=sfraser
Attachment #53030 - Flags: superreview+
(Assignee)

Comment 21

16 years ago
fixed. patches checked in to:
mozilla/build/mac/IDE_Options.h
mozilla/config/mac/NGLayoutConfigInclude.h
ns/build/mac/IDE_Options.h
Status: ASSIGNED → RESOLVED
Last Resolved: 16 years ago
Resolution: --- → FIXED

Comment 22

16 years ago
I'd suggest this go in on the 094 branch as well as the trunk.  Also on the 095 
branch if you can get drivers approval (and explain to them we get no debug info 
on OS X release builds if they balk at taking it)

Comment 23

16 years ago
Pls advise on how I can verify?
QA Contact: granrose → lchiang
(Reporter)

Comment 24

16 years ago
To verify: make the build crash somehow(?), and look at the crash reporter 
report. It should give function names in the report, rather than just hex values.

Comment 25

16 years ago
To clarify how one gets a CrashReporter report, assuming Mac OS X 10.1, run the 
Console application (found in Applications/Utilities).  Select the Preferences... 
item from the Console menu, click the Crashes tab in the prefs dialog, enable the 
checkboxes for Log crash information & Automatically display crash logs.

Comment 26

16 years ago
http://bugzilla.mozilla.org/show_bug.cgi?id=102727 is a way to crash on OS X.  
Petersen - I don't have OS X.1 (w/crash reporter installed). Can you verify the
crash on OSX branch builds to make sure the stack trace shows function names
instead of hex values?  If so, then this bug report is verified.
QA Contact: lchiang → petersen

Comment 27

16 years ago
Created attachment 54328 [details]
Crashreport of 0.9.5 on MacOSX

Comment 28

16 years ago
I have verified this is fixed in the 094 branch 2001-10-19 builds.  Used bug
101868 to generate the crash.  The log file shows the function names.

Still need to verify on trunk so will leave this bug not yet marked verified.
Keywords: vtrunk

Comment 29

16 years ago
Created attachment 54355 [details]
crash log from 094 branch

Updated

16 years ago
No longer blocks: 102998

Updated

16 years ago
Blocks: 102998
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.