Closed
Bug 214152
Opened 21 years ago
Closed 21 years ago
Crash (Sys3175) when opening URL as bookmarklet
Categories
(SeaMonkey :: General, defect)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: amai, Assigned: mkaply)
References
()
Details
(Keywords: crash, fixed1.4.1)
Attachments
(2 files, 1 obsolete file)
1.76 KB,
patch
|
jhpedemonte
:
review+
bzbarsky
:
superreview+
|
Details | Diff | Splinter Review |
1.02 KB,
patch
|
jhpedemonte
:
review+
mkaply
:
superreview+
mkaply
:
approval1.6b+
|
Details | Diff | Splinter Review |
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.5a) Gecko/20030718 Build Identifier: http://ftp.mozilla.org/pub/mozilla/releases/mozilla1.4/mozilla-os2-vacpp-1.4.zip I drag that link in the toolbar and click on it. Same if the link is to the same HTML source but on a local file. Reproducible: Always Steps to Reproduce: 1. 2. 3. Expected Results: Not crash, but display passwords in forms on current page
Sorry for omitting some information on the initial report. I can supply the exact crash info later today.
Version: Trunk → 1.4 Branch
Comment 2•21 years ago
|
||
---> Browser-General until we can get a stack trace
Assignee: rogerl → general
Component: JavaScript Engine → Browser-General
Keywords: crash
QA Contact: pschwartau → general
Summary: Sys3175 when opening URL as bookmarklet → Crash (Sys3175) when opening URL as bookmarklet
Assignee | ||
Comment 3•21 years ago
|
||
Taking
Assignee: general → mkaply
Status: UNCONFIRMED → NEW
Ever confirmed: true
Assignee | ||
Comment 4•21 years ago
|
||
The actual bug here appears to be that UniUconvFromUcs is failing with a long string. However, the NS_NewLocalFile API should be smart enough to handle this failing case, and not just return NS_OK all the time.
Assignee | ||
Comment 5•21 years ago
|
||
I took the linux code which is a little better and also put a patch for Windows.
Assignee | ||
Updated•21 years ago
|
Attachment #128716 -
Attachment is obsolete: true
Assignee | ||
Updated•21 years ago
|
Attachment #128717 -
Flags: review?(pedemont)
Ok, looks like a good part of the work already done :-) --> Thanks! Just for completness the precise crash info: Mozilla/5.0 (OS/2; U; Warp 4.5; en-US; rv:1.4) Gecko/20030724 07-28-2003 20:42:25 SYS3175 PID 002a TID 0001 Slot 0069 E:\APPS\MOZILLA\PM\BIN\MOZILLA.EXE c0000005 1def5d25 P1=00000001 P2=00000000 P3=XXXXXXXX P4=XXXXXXXX EAX=00000000 EBX=00000000 ECX=0014e604 EDX=0014e600 ESI=00000001 EDI=0014e600 DS=0053 DSACC=d0f3 DSLIM=3fffffff ES=0053 ESACC=d0f3 ESLIM=3fffffff FS=150b FSACC=00f3 FSLIM=00000030 GS=08b3 GSACC=10f3 GSLIM=00003fff CS:EIP=005b:1def5d25 CSACC=d0df CSLIM=3fffffff SS:ESP=0053:0014e5b4 SSACC=d0f3 SSLIM=3fffffff EBP=0014e6ac FLG=00012212 NECKO.DLL 0001:00005d25
Updated•21 years ago
|
Attachment #128717 -
Flags: review?(pedemont) → review+
Assignee | ||
Updated•21 years ago
|
Attachment #128717 -
Flags: superreview?(bzbarsky)
Updated•21 years ago
|
Severity: normal → critical
Comment 7•21 years ago
|
||
Comment on attachment 128717 [details] [diff] [review] Better fix sr=me, though you may want to also set the out param to null on the error return....
Attachment #128717 -
Flags: superreview?(bzbarsky) → superreview+
Assignee | ||
Comment 8•21 years ago
|
||
I checked in this fix, which does fix the problem. LEaving open because on OS/2 there is a bigger issue - why does the unicode conversion fail in such a strange way?
Assignee | ||
Comment 9•21 years ago
|
||
Put in 1.4 as well.
Updated•21 years ago
|
Keywords: fixed1.4.1
Assignee | ||
Comment 10•21 years ago
|
||
The previous fix fixed the bad NS_NewLocalFile stuff. There is a secondary OS/2 bug we need to workaround as well. Basically the UniUconv APIs require that the output buffer be at least as big as the input buffer or they fail immediately with an undocumented API. So rather than always passing in CCHMAXPATH, we pass in the size of inputLen.
Assignee | ||
Updated•21 years ago
|
Attachment #135932 -
Flags: review?(pedemont)
Comment 11•21 years ago
|
||
Comment on attachment 135932 [details] [diff] [review] More fix Add a better explanation to the comments on why resultLen must be greater than inputLen. Otherwise, r=
Attachment #135932 -
Flags: review?(pedemont) → review+
Assignee | ||
Comment 12•21 years ago
|
||
Comment on attachment 135932 [details] [diff] [review] More fix sr=blizzard (platform specific) a=mkaply for 1.6b
Attachment #135932 -
Flags: superreview+
Attachment #135932 -
Flags: approval1.6b+
Assignee | ||
Comment 13•21 years ago
|
||
This bug is fully fixed.
Status: NEW → RESOLVED
Closed: 21 years ago
Resolution: --- → FIXED
Updated•20 years ago
|
Product: Browser → Seamonkey
You need to log in
before you can comment on or make changes to this bug.
Description
•