You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@openoffice.apache.org by bu...@apache.org on 2012/07/03 10:01:28 UTC
[Bug 120165] New: [From Symphony]Impress crashed when play screen
show with sample file
https://issues.apache.org/ooo/show_bug.cgi?id=120165
Priority: P3
Bug ID: 120165
Assignee: ooo-issues@incubator.apache.org
Summary: [From Symphony]Impress crashed when play screen show
with sample file
Severity: critical
Issue Type: DEFECT
Classification: Application
OS: All
Reporter: yanji.yj@gmail.com
Hardware: All
Status: CONFIRMED
Version: AOO 3.4.0
Component: editing
Product: presentation
Created attachment 78561
--> https://issues.apache.org/ooo/attachment.cgi?id=78561&action=edit
sample
Build: AOO 3.4
Open attached sample file, then play screen show
Defect: Application crashed
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #14 from Armin Le Grand <Ar...@me.com> ---
ALG->pengyunquan: When the mask has not the same pixel size, it can anyways not
fit the content; the mask will anyways been scaled when painting with VCL (not
truncated as in the patch), so scaling it at BitmapEx construction time makes
no difference for visualization, except that the visualisation does the scaling
at runtime.
Changing the saved BitmapEx: Yes, it will be saved changed, but I think this is
acceptable for such bitmaps. As explained above, VCL will scale the Mask/Alpha
at output to the same output pixel size anyways. All pixel formats I know which
support Alpha/Mask have something like RGBA anyways in file formats, thus a
unequal SizePixel is not even possible/allowed since one alpha value is bound
to one pixel.
I also do not know how that BitmapEx was created, but as the existing asserts
show it is not intended and leads to many exceptions in the whole office. In
many places are asserts to warn in this case (as in the task case there were
already warnings), but no handling implemented. I have not designed VCL, but
this looks like purpose, too.
If there should really be a case where this is problematic we can react on it
when it comes up.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
pengyunquan <yu...@cn.ibm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #78748| |review?
Flags| |
--- Comment #3 from pengyunquan <yu...@cn.ibm.com> ---
Created attachment 78748
--> https://issues.apache.org/ooo/attachment.cgi?id=78748&action=edit
Fix patch for Bug 120165
Update fix patch
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
louqle <lo...@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |CLOSED
--- Comment #16 from louqle <lo...@gmail.com> ---
Verified on xp, Ubuntu, win7,mac,win7-64,Vista,suse,redhat, SLED 11 SP1 64 bit,
all pass, close this bug
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #13 from pengyunquan <yu...@cn.ibm.com> ---
If we scale the mask bitmap, we changed the saved BitmapEx object. I know it is
an abnormal BitmapEx object, and we don't know how this abnormal object is
generated. But I think we'd better keep the abnormal BitmapEx object to avoid
risk of incorrect display result.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #9 from Andre <aw...@googlemail.com> ---
Then go ahead and check in your changes.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
louqle <lo...@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |louqingle@gmail.com
--- Comment #15 from louqle <lo...@gmail.com> ---
verified on XP SP3 and Ubuntu 12.04 against trunk build 1374181
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #8 from Armin Le Grand <Ar...@me.com> ---
ALG: It replaces the original one. It fixes the general problem root by
adapting the mask/alpha bitmap at BitmapEx construction time if not the same
pixel size (there was an assert there already indicating that this is not
wanted).
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
Armin Le Grand <Ar...@me.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|CONFIRMED |ACCEPTED
CC| |Armin.Le.Grand@me.com
--- Comment #5 from Armin Le Grand <Ar...@me.com> ---
ALG: I have even more questions to this crash and fix.
- Why only fix for Mask, BitmapEx can also have Alpha with the same problem?
- Why not fix in a way that alpha/mask gets scaled and not just truncated?
- The BitmapEx constructor already asserts unequal sizes, this imples that
unequal sizes are not wanted from design. Why not fix there?
I will propose an other xix/patch.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
pengyunquan <yu...@cn.ibm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #78747| |review?
Flags| |
--- Comment #1 from pengyunquan <yu...@cn.ibm.com> ---
Created attachment 78747
--> https://issues.apache.org/ooo/attachment.cgi?id=78747&action=edit
Fix patch for Bug 120165
The sample file contains .svm files which contains bitmapex whose mask is not
the same size as its bitmap. And this cause access out of boundary in
GraphicManager::ImplCreateScaled
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
Armin Le Grand <Ar...@me.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ACCEPTED |RESOLVED
Resolution|--- |FIXED
--- Comment #11 from Armin Le Grand <Ar...@me.com> ---
ALG: Okay, comitted as revision 1366405, done.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #10 from SVN Robot <sv...@dev.null.org> ---
"alg" committed SVN revision 1366405 into trunk:
#120165# Adapt Mask/Alpha at BitmapEx construction when size differs from
bas...
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
Andre <aw...@googlemail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |awf.aoo@googlemail.com
--- Comment #4 from Andre <aw...@googlemail.com> ---
@pengyunquan: I have two questions regarding the fix:
1. Instead of introducing two new variables nDstWV and nDstHV, would it be
possible to modify and reuse the existing ones nDstW and nDstH?
2. The use pMapLX and pMapLY in lines 1251 and 1252 indicates that their values
are sorted. Is that true?
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
pengyunquan <yu...@cn.ibm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #78748|0 |1
is patch| |
Attachment #78748|application/octet-stream |text/plain
mime type| |
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
pengyunquan <yu...@cn.ibm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #78747|0 |1
is obsolete| |
Attachment #78747|review? |
Flags| |
--- Comment #2 from pengyunquan <yu...@cn.ibm.com> ---
Comment on attachment 78747
--> https://issues.apache.org/ooo/attachment.cgi?id=78747
Fix patch for Bug 120165
update a patch with full source file path later
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #7 from Andre <aw...@googlemail.com> ---
@ALG: Is this an additional patch or does it replace the previous one?
@pengyunquan: Is the second patch OK for you?
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
Shenfeng Liu <li...@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |liushenf@gmail.com
Target Milestone|--- |AOO 3.5.0
--- Comment #17 from Shenfeng Liu <li...@gmail.com> ---
set Target Milestone to AOO 3.5.0 for PM purpose.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
--- Comment #6 from Armin Le Grand <Ar...@me.com> ---
Created attachment 78754
--> https://issues.apache.org/ooo/attachment.cgi?id=78754&action=edit
Fix by correcting Mask/Alpha at BitmapEx construction time
ALG: Patch added which corrects given Alpha/Mask at BitmapEx construction time,
thus ensuring unequal pixel sizes will not happen. This makes all code based on
BitmapEx pixel magic involving Alpha/Mask more simple and safe.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 120165] [From Symphony]Impress crashed when play screen show
with sample file
Posted by bu...@apache.org.
https://issues.apache.org/ooo/show_bug.cgi?id=120165
pengyunquan <yu...@cn.ibm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |yunquanp@cn.ibm.com
--- Comment #12 from pengyunquan <yu...@cn.ibm.com> ---
I have one question: If we scale the mask bitmap, is the mask bitmap still a
valid mask? As we know even a pixel offset of mask bitmap will get quite
different mask result.
--
You are receiving this mail because:
You are the assignee for the bug.