Re: [RFC2] new findPointCoords implementation (and bug triage(sp?)) (fwd)


Subject: Re: [RFC2] new findPointCoords implementation (and bug triage(sp?)) (fwd)
From: sam th (sam@uchicago.edu)
Date: Sun Jun 04 2000 - 03:22:16 CDT


---------- Forwarded message ----------
Date: Sat, 3 Jun 2000 05:48:28 -0500 (CDT)
From: sam th <sam@uchicago.edu>
To: Jesper Skov <jskov@redhat.com>
Cc: Martin Sevior <msevior@mccubbin.ph.unimelb.edu.au>,
     abiword-dev@abisource.com
Subject: Re: [RFC2] new findPointCoords implementation (and bug triage(sp?))

On Sat, 3 Jun 2000, Jesper Skov wrote:

> It fixes the below Bugs (i.e., I could reproduce with a pre-patched
> build, but not with the pathed build):
> 852, 851, 841, 823, 449, 448,

I can confirm 852, 851, 841, 823.

In trying to test 841 I discovered that inserting a page break in a blank
document causes the following assert, and then a segfault.
**** (1) Assert ****
**** (1) NULL != pRun at fl_BlockLayout.cpp:964 ****
**** (1) Continue ? (y/n) [y] : y
This is caused by your patch, AFAIK. It does not happend if you have
already typed and deleted a page break, but only on a 1 page document.

The stack trace looks like:

#0 0x812131d in fl_BlockLayout::findPointCoords (this=0x8329758, iPos=3,
bEOL=0 '\000',
    x=@0xbffff25c, y=@0xbffff258, height=@0xbffff254) at
fl_BlockLayout.cpp:1026
#1 0x8116aef in FV_View::getCharFormat (this=0x8329970,
pProps=0xbffff2d8,
    bExpandStyles=1 '\001') at fv_View.cpp:1482
#2 0x811464b in FV_View::notifyListeners (this=0x8329970, hint=32) at
fv_View.cpp:331
#3 0x8123e14 in fl_BlockLayout::doclistener_insertSpan (this=0x8329758,
pcrs=0x82c7f48)
    at fl_BlockLayout.cpp:2266
#4 0x812a86a in fl_SectionLayout::bl_doclistener_insertSpan
(this=0x8329620,
    pBL=0x8329758, pcrs=0x82c7f48) at fl_SectionLayout.cpp:272
#5 0x81295d6 in fl_DocListener::change (this=0x8329560, sfh=0x8329758,
pcr=0x82c7f48)
    at fl_DocListener.cpp:310
#6 0x813d8ec in PD_Document::notifyListeners (this=0x8323908,
pfs=0x8328e00,
    pcr=0x82c7f48) at pd_Document.cpp:460
#7 0x814754c in pt_PieceTable::insertSpan (this=0x8324190, dpos=2,
p=0xbffff4f2,
    length=1) at pt_PT_InsertSpan.cpp:434
#8 0x813d2b7 in PD_Document::insertSpan (this=0x8323908, dpos=2,
p=0xbffff4f2, length=1,
    p_AttrProp=0x0) at pd_Document.cpp:265
#9 0x81160cb in FV_View::cmdCharInsert (this=0x8329970, text=0xbffff4f2,
count=1,
    bForce=0 '\000') at fv_View.cpp:1075
#10 0x80ce55b in ap_EditMethods::insertPageBreak (pAV_View=0x8329970)
    at ap_EditMethods.cpp:2410

Testing 449 caused the same assert, and then the same segfault.

Testing 448 also causes a segfault, with the following prior asserts.

**** (1) Assert ****
**** (1) m_iHeightLayoutUnits at fp_Line.cpp:391 ****
**** (1) Continue ? (y/n) [y] : y
y^H
**** (2) Assert ****
**** (2) m_iHeightLayoutUnits at fp_Line.cpp:391 ****
**** (2) Continue ? (y/n) [y] : y

**** (3) Assert ****
**** (3) m_iHeightLayoutUnits at fp_Line.cpp:391 ****
**** (3) Continue ? (y/n) [y] : y

**** (4) Assert ****
**** (4) pClosestRun at fp_Line.cpp:266 ****
**** (4) Continue ? (y/n) [y] : y

More debugging to follow.
                                     sam th
                                     sam@uchicago.edu
                                http://sam.rh.uchicago.edu
------------ Output from gpg ------------
gpg: Good signature from "Sam Tobin-Hochstadt <sytobinh@uchicago.edu>"



This archive was generated by hypermail 2b25 : Sun Jun 04 2000 - 03:24:15 CDT