Jesse Ruderman Reporter
	Description • 16 years ago

Re bug 50630 comment 51:

> > +    NS_ASSERTION(psd->mRightEdge >= psd->mLeftEdge,
> > +                 "We placed a float where there was no room!");
> > 
> > Are you sure you can't hit this with negative margins?
> 
> We don't hit it in reftests.

Famous last words.  This testcase hits it:

###!!! ASSERTION: We placed a float where there was no room!: 'psd->mX <= psd->mRightEdge', file /Users/jruderman/central/mozilla/layout/generic/nsLineLayout.cpp, line 344

	Robert O'Callahan (:roc) (email my personal email if necessary)
	Comment 1 • 16 years ago

Ah.

We should probably just take the assertion out, since in retrospect it's bound to fire when the float is the first thing on the line and the float is wider than the available width.

	Jesse Ruderman Reporter
	Comment 2 • 16 years ago

roc changed the assertion condition in bug 441259.  The testcase in this bug still hits the modified assertion:

###!!! ASSERTION: We placed a float where there was no room!: 'psd->mX - mTrimmableWidth <= psd->mRightEdge', file /Users/jruderman/central/mozilla/layout/generic/nsLineLayout.cpp, line 345

	Olli Pettay [:smaug][bugs@pettay.fi]
	Comment 3 • 16 years ago

Something has changed recently; nowadays running mochitest triggers this a lot.

	neil@parkwaycc.co.uk
	Comment 4 • 16 years ago

(In reply to comment #3)
> Something has changed recently; nowadays running mochitest triggers this a lot.
Just browsing to planet.mozilla.org triggers this assertion twice for me.

	Robert O'Callahan (:roc) (email my personal email if necessary)
	Comment 5 • 16 years ago

The new nsFloatManager is the most likely suspect...

	David Baron :dbaron: Assignee
	Comment 6 • 16 years ago

(In reply to comment #3)
> Something has changed recently; nowadays running mochitest triggers this a lot.

That should be filed as a bug separate from this one if it still exists after the landing of bug 472252.

	Mats Palmgren (inactive)
	Comment 7 • 16 years ago

The testcase in bug 435664 also triggers it; current trunk on Linux.

	David Baron :dbaron: Assignee
	Comment 8 • 16 years ago

I filed the separate bug as bug 476372.

	David Baron :dbaron: Assignee
	Comment 9 • 16 years ago

(In reply to comment #7)
> The testcase in bug 435664 also triggers it; current trunk on Linux.

Those assertions were prior to the float manager landing, so thus are not covered by bug 476372.

	Ben Turner (not reading bugmail, use the needinfo flag!)
	Comment 10 • 16 years ago

Hopefully this isn't just noise but dromaeo triggers this a lot too.

	Jesse Ruderman Reporter
	Comment 11 • 14 years ago

379217-2.xhtml hits this assertion too.

	Andreas Gal :gal
	Comment 12 • 14 years ago

This is really annoying in debug builds and makes debugging with fatal asserts borderline impossible. If this can't be fixed, we should switch the assert off and convert it to a silent FIXME with a bug #.

###!!! ASSERTION: We placed a float where there was no room!: 'psd->mX - mTrimmableWidth <= psd->mRightEdge', file ../../../layout/generic/nsLineLayout.cpp, line 345
###!!! ASSERTION: We placed a float where there was no room!: 'psd->mX - mTrimmableWidth <= psd->mRightEdge', file ../../../layout/generic/nsLineLayout.cpp, line 345

	David Baron :dbaron: Assignee
	Comment 14 • 11 years ago

https://hg.mozilla.org/integration/mozilla-inbound/rev/5ddb2542c961

(also see try push at https://tbpl.mozilla.org/?tree=Try&rev=657c8987d64d)

	Ed Morley [:emorley]
	Comment 15 • 11 years ago

https://hg.mozilla.org/mozilla-central/rev/5ddb2542c961