REGRESSION: Chrome PDF Viewer Can't view second and beyond page of pdf |
|||||||||||||||||||||
Issue descriptionVersion 52.0.2710.0 canary (64-bit) Win10 http://www.questrade.com/docs/default-source/downloads/faqs_tax.pdf I get only the first page, and then a vertical ellipsis that doesn't seem to do anything. Works OK on Stable Version 49.0.2623.112 m.
,
Apr 16 2016
Tried turning off all about:flags (was just a few MD-related ones), but no change. Variations 16e0dd70-3f4a17df b3888d8d-52a0c8a0 da89714-4ad60575 6345b824-3f4a17df 7c1bc906-f55a7974 c4455df5-abb3ea59 a5cb8590-3f4a17df f049a919-3f4a17df f1aba312-4ef7a373 775ebbd7-3f4a17df 31362330-3f4a17df f15c1c09-ca7d8d80 9ffc5535-301ba80c dd4da2fc-3f4a17df fd02e767-ca7d8d80 93731dca-3f4a17df 9e5c75f1-f5d7252d 2c3080ba-3f4a17df 64cbdfc2-ca7d8d80 7ea1191-82bd42b5 f79cb77b-3f4a17df 23a898eb-ca7d8d80 4ea303a6-f23d1dea d5b671a5-3f4a17df 4117e878-23622c9b 7aa46da5-4995b330 9736de91-3f4a17df 535ce4f1-55992af7 ad6d27cc-c6d02f41 ca314179-ca7d8d80 69bf80fa-f23d1dea a35118-3d47f4f4 867c4c68-3d47f4f4 5e3f6590-ca7d8d80 12a73824-3f4a17df d747916f-d747916f 6844d8aa-669a04e0 3ac60855-486e2a9c f296190c-5192b0cf 4442aae2-6e3b1976 ed1d377-e1cc0f14 75f0f0a0-a5822863 e2b18481-6e597ede e7e71889-e1cc0f14 b39ea213-d1372334 46567c16-3d47f4f4 b0dc61a1-ca7d8d80
,
Apr 18 2016
hmm I was able to repro this in the end, on a different machine. With that, I was able to bisect to: You are probably looking for a change made after 372412 (known good), but no later than 372427 (first known bad). CHANGELOG URL: https://chromium.googlesource.com/chromium/src/+log/64cd59d..ba874b5?pretty=fuller the only PDF-ish CL is https://codereview.chromium.org/1636413003 which doesn't make sense as that's quite an old CL, and I can't repro on stable 50.0.2661.75. Strange.
,
Apr 18 2016
oh yes actually I can repro on stable if I maximize the window so it's larger than the page, before loading the PDF file.
,
Apr 18 2016
It works fine if I --force-device-scale-factor=2 or 3. It doesn't work with --force-device-scale-factor=1. It also seems fine on my Mac at 2710 or 2711 which is a laptop and dsf=2. Something to do with how much visible area there is compared to the size of the first page I think.
,
Apr 18 2016
I can't repro with 52.0.2709.0 Canary via RDP. I tried maximizing my window and adding --force-device-scale-factor=1 but no luck. Is this my excuse for upgrading to a high DPI monitor? ;)
,
Apr 22 2016
It only repros for me at a particular height or above. It also only repros with certain files for some reason -- it doesn't repro if I just save a page as PDF and open that. In my case it occurs when window.innerHeight is 1155px or above, regardless of window width. The div#sizer is hard-coded at 1164px if that means anything. There are three vertical dots below the page which are part of the plugin itself. 50.0.2661.75 and 52.0.2714.0 on Windows.
,
Apr 23 2016
I'm also having this issue on 51.0.2704.19 dev-m (64-bit) on Windows 8.1 and 51.0.2700.0 dev (64-bit) on OSX 10.10.5. I can view all the pages if the window is smaller than the page size when I open/refresh the page.
,
Apr 23 2016
,
Apr 23 2016
Any idea if this regressed? The sizer element height calculation is from r342584.
,
Apr 23 2016
I bisected in #3
,
Apr 25 2016
Whoops, missed that. I can repro on Linux if I just run a local VNC server at a very high resolution.
,
Apr 25 2016
And it is indeed r372415 that caused this, so I'll try to figure it out. I guess with the VNC server repro, I don't have an excuse to get myself a new monitor.
,
Apr 26 2016
So we switched to FPDFAvail_GetDocument() for successfully loaded documents that are linearized, which solved another bug, but for this document, FPDF_GetPageSizeByIndex() is failing early on for pages 2-4. Thus their sizes are 0x0, and they show up as those funny squares.
,
Apr 29 2016
,
May 2 2016
I have no way to check if it's the same bug or not, but this PDF http://academique.epfl.ch/calendrier_academique/EPFL-Academic-Calendar.pdf has the same problem. Interestingly, if I rotate the page I can scroll normally. If I rotate it 4 times (i.e. 360 degrees), I can still scroll normally.
,
May 2 2016
Another file that triggers this issue is http://www.vldb.org/conf/2001/P541.pdf if that helps you find the cause. Chrome stable 50.0.2661.94 (64-bit) on Debian unstable.
,
May 2 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/21699f821ea44f1d34adc94cc0f8abbfb22c1525 commit 21699f821ea44f1d34adc94cc0f8abbfb22c1525 Author: thestig <thestig@chromium.org> Date: Mon May 02 18:48:30 2016 Get the correct page size for linearized PDFs. BUG= 604146 Review-Url: https://codereview.chromium.org/1939693002 Cr-Commit-Position: refs/heads/master@{#391011} [modify] https://crrev.com/21699f821ea44f1d34adc94cc0f8abbfb22c1525/pdf/pdfium/pdfium_engine.cc
,
May 2 2016
Let's see if the fix works on Canary before merging to M51.
,
May 3 2016
52.0.2723.0 (current Canary) and newer should have the fix. Please give it a try and see if linearized PDFs are displaying properly again.
,
May 3 2016
Yeah, lg for me.
,
May 3 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/3fc57cf9f244272e6ab32b2db080fbcf4349d95d commit 3fc57cf9f244272e6ab32b2db080fbcf4349d95d Author: weili <weili@chromium.org> Date: Tue May 03 20:24:19 2016 Revert of Get the correct page size for linearized PDFs. (patchset #1 id:1 of https://codereview.chromium.org/1939693002/ ) Reason for revert: Speculative revert due to high crash volume in Chromium. Original issue's description: > Get the correct page size for linearized PDFs. > > BUG= 604146 > > Committed: https://crrev.com/21699f821ea44f1d34adc94cc0f8abbfb22c1525 > Cr-Commit-Position: refs/heads/master@{#391011} TBR=spelchat@chromium.org,thestig@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. BUG= 604146 ,608698 Review-Url: https://codereview.chromium.org/1942253003 Cr-Commit-Position: refs/heads/master@{#391342} [modify] https://crrev.com/3fc57cf9f244272e6ab32b2db080fbcf4349d95d/pdf/pdfium/pdfium_engine.cc
,
May 3 2016
lg for me on canary as well.
,
May 3 2016
Reverted to fix issues 608778
,
May 3 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/2aeafdc83d22cd63fe89ddbce7fbc90bd537b9e1 commit 2aeafdc83d22cd63fe89ddbce7fbc90bd537b9e1 Author: Vitaly Buka <vitalybuka@chromium.org> Date: Tue May 03 21:40:31 2016 Merge: Revert of Get the correct page size for linearized PDFs. (patchset #1 id:1 of https://codereview.chromium.org/1939693002/ ) Reason for revert: Speculative revert due to high crash volume in Chromium. BUG=608778 Original issue's description: > Get the correct page size for linearized PDFs. > > BUG= 604146 > > Committed: https://crrev.com/21699f821ea44f1d34adc94cc0f8abbfb22c1525 > Cr-Commit-Position: refs/heads/master@{#391011} TBR=spelchat@chromium.org,thestig@chromium.org BUG= 604146 ,608698 Review-Url: https://codereview.chromium.org/1942253003 Cr-Commit-Position: refs/heads/master@{#391342} (cherry picked from commit 3fc57cf9f244272e6ab32b2db080fbcf4349d95d) Review URL: https://codereview.chromium.org/1950693002 . Cr-Commit-Position: refs/branch-heads/2723@{#2} Cr-Branched-From: 5e216d9b62275101c30f4a5cb2174bb68970ef3c-refs/heads/master@{#391139} [modify] https://crrev.com/2aeafdc83d22cd63fe89ddbce7fbc90bd537b9e1/pdf/pdfium/pdfium_engine.cc
,
May 5 2016
Issue 609322 has been merged into this issue.
,
May 5 2016
,
May 9 2016
A friendly reminder that M51 Stable is launching soon! Your bug is labelled as Stable ReleaseBlock, pls make sure to land the fix and get it merged into the release branch by May 17. All changes MUST be merged into the release branch by 5pm on May 20 to make into the desktop Stable final build cut. Thanks!
,
May 11 2016
vitalybuka@ : Could you please providing the repro steps to verify from Test Engg team if at all there is specific steps to follow. Unable to reproduce the issue with or witout "--force-device-scale-factor=1" on 52.0.2730.0.
,
May 11 2016
On OS X, I can repro this on 52.0.2730.0 following these steps: 1) Make the browser window tall (1575px tall on my machine). 2) Open https://www.csimfunds.com/public/file/P-6970550/SLS37226_38_Fundamental_Index.pdf 3) Pages after the first page are tiny blobs
,
May 12 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/8ad3c81f1ec4e67a71b17f7d97165c68a3c87d46 commit 8ad3c81f1ec4e67a71b17f7d97165c68a3c87d46 Author: thestig <thestig@chromium.org> Date: Thu May 12 01:11:49 2016 Get the correct page size for linearized PDFs. This is a reland of r391011, which was reverted because the new code path hit a PDFium bug. BUG= 604146 Review-Url: https://codereview.chromium.org/1975463002 Cr-Commit-Position: refs/heads/master@{#393128} [modify] https://crrev.com/8ad3c81f1ec4e67a71b17f7d97165c68a3c87d46/pdf/pdfium/pdfium_engine.cc
,
May 12 2016
Hopefully the fix sticks this time. It will be in 52.0.2734.0 and newer.
,
May 13 2016
Issue 609119 has been merged into this issue.
,
May 13 2016
Issue 607574 has been merged into this issue.
,
May 13 2016
I've gotten positive feedback on Canary from multiple sources, so requesting a merge to M51: - Need to merge to PDFium's M51 branch: https://pdfium.googlesource.com/pdfium/+/35c2100a5f6a466635bf99b4e7117d23aeb54d2c - Then roll DEPS for PDFium in Chromium's M51 branch - And merge https://chromium.googlesource.com/chromium/src.git/+/8ad3c81f1ec4e67a71b17f7d97165c68a3c87d46
,
May 13 2016
Your change meets the bar and is auto-approved for M51 (branch: 2704)
,
May 13 2016
Works for me on 52.0.2735.0. Thanks!
,
May 13 2016
Oh you people and your fancy monitors.
,
May 13 2016
The following revision refers to this bug: http://goto.ext.google.com/viewvc/chrome-internal?view=rev&revision=87843 ------------------------------------------------------------------ r87843 | thestig@google.com | 2016-05-13T21:51:15.126242Z -----------------------------------------------------------------
,
May 13 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/51ea00a747d4bee74bfeb0e52116274f617649c9 commit 51ea00a747d4bee74bfeb0e52116274f617649c9 Author: Lei Zhang <thestig@chromium.org> Date: Fri May 13 21:55:40 2016 Merge to M51: Get the correct page size for linearized PDFs. This is a reland of r391011, which was reverted because the new code path hit a PDFium bug. BUG= 604146 Review-Url: https://codereview.chromium.org/1975463002 Cr-Commit-Position: refs/heads/master@{#393128} (cherry picked from commit 8ad3c81f1ec4e67a71b17f7d97165c68a3c87d46) Review URL: https://codereview.chromium.org/1974833006 . Cr-Commit-Position: refs/branch-heads/2704@{#546} Cr-Branched-From: 6e53600def8f60d8c632fadc70d7c1939ccea347-refs/heads/master@{#386251} [modify] https://crrev.com/51ea00a747d4bee74bfeb0e52116274f617649c9/pdf/pdfium/pdfium_engine.cc
,
May 13 2016
,
May 18 2016
Tested the issue on Windows 7, 10, Mac 10.11.5, Ubuntu 14.04 using 51.0.2704.54. Able to see all pages of PDF of comment #0, comment #30 without any font tiny blobs. Please find attached screencast. Marking it as TE-Verified.
,
May 19 2016
Issue 609936 has been merged into this issue.
,
May 19 2016
Issue 613308 has been merged into this issue.
,
May 19 2016
,
May 24 2016
|
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by wfh@chromium.org
, Apr 16 2016