This is a spin off bug for Bug 484932. Steps to reproduce: Download the testcase https://bugzilla.mozilla.org/attachment.cgi?id=369466 Untar it. Open test.html. Print it. In terminal I got some assertions. It's reproducible on all platforms. ###!!! ASSERTION: Unexpected containers: 'SameCOMIdentity(debugDocContainer, debugDocShell)', file ../../../layout/base/nsDocumentViewer.cpp, line 2083 nsBlockReflowContext: Frameset(frameset)(1)@0xd83198 metrics=38960,1073741824! ###!!! ASSERTION: Unexpected containers: 'SameCOMIdentity(debugDocContainer, debugDocShell)', file ../../../layout/base/nsDocumentViewer.cpp, line 2083 ###!!! ASSERTION: There must always be an XMost PO!: 'smallestPO', file ../../../layout/printing/nsPrintEngine.cpp, line 1659 ###!!! ASSERTION: There must always be an XMost PO!: 'smallestPO', file ../../../layout/printing/nsPrintEngine.cpp, line 1704 Back out Bug 467881 doesn't resolve it.

I'm seeing this assertion by print previewing a page with an iframe.

On Linux, with the testcase in comment 1, I get: ###!!! ASSERTION: Unexpected containers: 'SameCOMIdentity(debugDocContainer, debugDocShell)', file layout/base/nsDocumentViewer.cpp, line 2134 ###!!! ASSERTION: no page sequence frame: 'nsnull != pageSequence', file layout/printing/nsPrintEngine.cpp, line 2282

The reduced testcase on Bug 652929 -- attachment 531449 [details] -- hits this bug's "SameCOMIdentity" assertion failure. However, it only hits it during a Print operation -- not during Print Preview. (As noted on that bug, the original issue there was also specific to actual-printing as opposed to print preview.) That testcase doesn't have any iframes -- just divs with the overflow, float, & clear properties set. (In contrast, I double-checked that Jesse's testcase here reproduces the assertion under both Print Preview and Print -- it does.)

FWIW, Bug 468568 Part 0.8 removed the "Unexpected containers" assertion: https://hg.mozilla.org/mozilla-central/rev/d7a64a002069

Cool. I get no assertions when printing the testcase mentioned in comment 4 now. However -- I do still see these assertions (some of which were mentioned in comment 0), when printing/print-previewing "test.html" in Jesse's attached testcase: { ###!!! ASSERTION: Shouldn't have unconstrained stuff here thanks to ComputeAutoSize: 'NS_INTRINSICSIZE != aReflowState.ComputedHeight()', file /mozilla-central/layout/generic/nsLeafFrame.cpp, line 78 nsBlockReflowContext: Frameset(frameset)(1)@0x7f7270897168 metrics=40320,1073741824! ###!!! ASSERTION: There must always be an XMost PO!: 'smallestPO', file /mozilla-central/layout/printing/nsPrintEngine.cpp, line 1704 ###!!! ASSERTION: There must always be an XMost PO!: 'smallestPO', file /mozilla-central/layout/printing/nsPrintEngine.cpp, line 1724 } It looks like the "Xmost" assertions are tracked in bug 410166, but there's no bug filed on the NS_INTRINSICSIZE assertion. Maybe we should morph this bug to cover that? Alternately, we can spin off a new bug and resolve this one as WFM.

(In reply to Daniel Holbert [:dholbert] (vacation through 8/31) from comment #6) > It looks like the "Xmost" assertions are tracked in bug 410166, but there's > no bug filed on the NS_INTRINSICSIZE assertion. Maybe we should morph this > bug to cover that? Alternately, we can spin off a new bug and resolve this > one as WFM. It's tracked in bug 441680, isn't it? Anyway, spinning off new bugs is better because my tools will know to stop ignoring this bug's assertion.