|
|
Created:
4 years, 6 months ago by ccameron Modified:
3 years, 10 months ago Reviewers:
dcdoherty, Ken Russell (switch to Gerrit), dub909, elfrank.ab, peter.saal, craiig, tsherif, robert.megone, aaron.whyte, Zhenyao Mo, vasiliskrikonis, peter, pallzoltan, ungoldman, arnaud.godard, sven, ktedore, emilevictor, tirkes CC:
chromium-reviews, piman+watch_chromium.org Base URL:
https://chromium.googlesource.com/chromium/src.git@master Target Ref:
refs/pending/heads/master Project:
chromium Visibility:
Public. |
DescriptionBlacklist HD3000 on Mac
HD3000 is responsible for 30% of our GPU process crashes on Mac.
BUG=592130
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel
Committed: https://crrev.com/0b113ce33d9d553b105fa68a2b2b554a93de43e9
Cr-Commit-Position: refs/heads/master@{#400262}
Patch Set 1 #
Messages
Total messages: 47 (7 generated)
Description was changed from ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 ========== to ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ==========
ccameron@chromium.org changed reviewers: + zmo@chromium.org
HD4000 is also on the chopping block, but this just does HD3000 for now. It holds the dubious honor of being responsible for 30% of our GPU crashes.
Description was changed from ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ========== to ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ==========
zmo@chromium.org changed reviewers: + kbr@chromium.org
On 2016/06/16 01:07:33, ccameron wrote: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. Maybe just turning off rasterization rather than everything? These GPUs should run WebGL OK. We don't want the situation where Firefox and Safari can have WebGL but Chrome doesn't. Adding Ken for his opinion.
The crashes on these GPUs were happening before GPU rasterization was turned on. The HD 3000 in particular is really old, and has been problematic. For these reasons I think it's OK to blacklist all GPU functionality on it, including WebGL. Users can manually specify --ignore-gpu-blacklist if they want to try content on it. LGTM from my standpoint.
OK, lgtm then
The CQ bit was checked by ccameron@chromium.org
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/patch-status/2076443002/1
Message was sent while issue was closed.
Description was changed from ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ========== to ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ==========
Message was sent while issue was closed.
Committed patchset #1 (id:1)
Message was sent while issue was closed.
Description was changed from ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel ========== to ========== Blacklist HD3000 on Mac HD3000 is responsible for 30% of our GPU process crashes on Mac. BUG=592130 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel Committed: https://crrev.com/0b113ce33d9d553b105fa68a2b2b554a93de43e9 Cr-Commit-Position: refs/heads/master@{#400262} ==========
Message was sent while issue was closed.
Patchset 1 (id:??) landed as https://crrev.com/0b113ce33d9d553b105fa68a2b2b554a93de43e9 Cr-Commit-Position: refs/heads/master@{#400262}
Message was sent while issue was closed.
I've used WebGL on this computer for over a year with no problem, until now. Now I'm in "the situation where Firefox and Safari can have WebGL but Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on my 2011-era MBP. -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
You guys are way too aggressive on this stuff. On Tuesday, August 2, 2016 at 7:19:38 PM UTC-5, aaron...@gmail.com wrote: > > I've used WebGL on this computer for over a year with no problem, until > now. Now I'm in "the situation where Firefox and Safari can have WebGL but > Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on > my 2011-era MBP. -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
I have the good fortune of owning an early 2011 MBP with an Intel HD 3000. Never had any problems with WebGL until y'all decided to disable it. My favorite part was the link in `chrome://gpu`: > Intel HD 3000 driver crashes frequently on Mac: 592130 <http://crbug.com/592130> > Disabled Features: all Following the link, I get the following friendly message: > You do not have permission to view the requested page. > > Reason: User is not allowed to view this issue A couple hours later I found this thread. Guess I'm manually setting --ignore-gpu-blacklist now. Not exactly the ideal user experience. On Tuesday, August 2, 2016 at 5:19:38 PM UTC-7, aaron...@gmail.com wrote: > > I've used WebGL on this computer for over a year with no problem, until > now. Now I'm in "the situation where Firefox and Safari can have WebGL but > Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on > my 2011-era MBP. -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 on too aggressive. We have customers who have used our WebGL app for over a year with no issue. Now they have an issue because of the blacklisting and we look stupid for taking them down this path. The solution was to use Firefox. On Tuesday, September 20, 2016 at 11:43:56 AM UTC-6, ungo...@gmail.com wrote: > > I have the good fortune of owning an early 2011 MBP with an Intel HD 3000. > > Never had any problems with WebGL until y'all decided to disable it. > > My favorite part was the link in `chrome://gpu`: > > > Intel HD 3000 driver crashes frequently on Mac: 592130 > <http://crbug.com/592130> > > Disabled Features: all > > Following the link, I get the following friendly message: > > > You do not have permission to view the requested page. > > > > Reason: User is not allowed to view this issue > > A couple hours later I found this thread. Guess I'm manually setting --ignore-gpu-blacklist > now. Not exactly the ideal user experience. > > On Tuesday, August 2, 2016 at 5:19:38 PM UTC-7, aaron...@gmail.com wrote: >> >> I've used WebGL on this computer for over a year with no problem, until >> now. Now I'm in "the situation where Firefox and Safari can have WebGL but >> Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on >> my 2011-era MBP. > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
agree. +1 on too aggressive. On Friday, September 23, 2016 at 2:20:53 PM UTC-4, peter...@gmail.com wrote: > > +1 on too aggressive. > > We have customers who have used our WebGL app for over a year with no > issue. Now they have an issue because of the blacklisting and we look > stupid for taking them down this path. The solution was to use Firefox. > > On Tuesday, September 20, 2016 at 11:43:56 AM UTC-6, ungo...@gmail.com > wrote: >> >> I have the good fortune of owning an early 2011 MBP with an Intel HD 3000. >> >> Never had any problems with WebGL until y'all decided to disable it. >> >> My favorite part was the link in `chrome://gpu`: >> >> > Intel HD 3000 driver crashes frequently on Mac: 592130 >> <http://crbug.com/592130> >> > Disabled Features: all >> >> Following the link, I get the following friendly message: >> >> > You do not have permission to view the requested page. >> > >> > Reason: User is not allowed to view this issue >> >> A couple hours later I found this thread. Guess I'm manually setting --ignore-gpu-blacklist >> now. Not exactly the ideal user experience. >> >> On Tuesday, August 2, 2016 at 5:19:38 PM UTC-7, aaron...@gmail.com wrote: >>> >>> I've used WebGL on this computer for over a year with no problem, until >>> now. Now I'm in "the situation where Firefox and Safari can have WebGL but >>> Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on >>> my 2011-era MBP. >> >> -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
On 2016/06/16 17:25:46, Ken Russell wrote: > The crashes on these GPUs were happening before GPU rasterization was turned on. > The HD 3000 in particular is really old, and has been problematic. For these > reasons I think it's OK to blacklist all GPU functionality on it, including > WebGL. Users can manually specify --ignore-gpu-blacklist if they want to try > content on it. > > LGTM from my standpoint. HD 3000 isn't _that_ old? I'm on a 2011 macbook air and thankfully found this thread and the flag to enable. I'm guessing most wont.
Message was sent while issue was closed.
+1, no issues until recently. Now I'm having to restart the browser upwards of three times a day. On Saturday, 24 September 2016 20:55:52 UTC+1, pe...@kokowa.co wrote: > > agree. +1 on too aggressive. > > On Friday, September 23, 2016 at 2:20:53 PM UTC-4, peter...@gmail.com > wrote: >> >> +1 on too aggressive. >> >> We have customers who have used our WebGL app for over a year with no >> issue. Now they have an issue because of the blacklisting and we look >> stupid for taking them down this path. The solution was to use Firefox. >> >> On Tuesday, September 20, 2016 at 11:43:56 AM UTC-6, ungo...@gmail.com >> wrote: >>> >>> I have the good fortune of owning an early 2011 MBP with an Intel HD >>> 3000. >>> >>> Never had any problems with WebGL until y'all decided to disable it. >>> >>> My favorite part was the link in `chrome://gpu`: >>> >>> > Intel HD 3000 driver crashes frequently on Mac: 592130 >>> <http://crbug.com/592130> >>> > Disabled Features: all >>> >>> Following the link, I get the following friendly message: >>> >>> > You do not have permission to view the requested page. >>> > >>> > Reason: User is not allowed to view this issue >>> >>> A couple hours later I found this thread. Guess I'm manually setting --ignore-gpu-blacklist >>> now. Not exactly the ideal user experience. >>> >>> On Tuesday, August 2, 2016 at 5:19:38 PM UTC-7, aaron...@gmail.com >>> wrote: >>>> >>>> I've used WebGL on this computer for over a year with no problem, until >>>> now. Now I'm in "the situation where Firefox and Safari can have WebGL but >>>> Chrome doesn't" that you wanted to avoid... I've got that Intel 3000 HD, on >>>> my 2011-era MBP. >>> >>> -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Same as above, 2011 MBP, never had real issues with my HD 3000 until few days back. What's worse, using ignore-gpu-blacklist results in WebGL crashing all the time now. On Thursday, September 29, 2016 at 10:44:44 AM UTC+3, robert...@gmail.com wrote: > > +1, no issues until recently. Now I'm having to restart the browser > upwards of three times a day. > > On Saturday, 24 September 2016 20:55:52 UTC+1, pe...@kokowa.co wrote: >> >> agree. +1 on too aggressive. >> >> On Friday, September 23, 2016 at 2:20:53 PM UTC-4, peter...@gmail.com >> wrote: >>> >>> +1 on too aggressive. >>> >>> We have customers who have used our WebGL app for over a year with no >>> issue. Now they have an issue because of the blacklisting and we look >>> stupid for taking them down this path. The solution was to use Firefox. >>> >>> On Tuesday, September 20, 2016 at 11:43:56 AM UTC-6, ungo...@gmail.com >>> wrote: >>>> >>>> I have the good fortune of owning an early 2011 MBP with an Intel HD >>>> 3000. >>>> >>>> Never had any problems with WebGL until y'all decided to disable it. >>>> >>>> My favorite part was the link in `chrome://gpu`: >>>> >>>> > Intel HD 3000 driver crashes frequently on Mac: 592130 >>>> <http://crbug.com/592130> >>>> > Disabled Features: all >>>> >>>> Following the link, I get the following friendly message: >>>> >>>> > You do not have permission to view the requested page. >>>> > >>>> > Reason: User is not allowed to view this issue >>>> >>>> A couple hours later I found this thread. Guess I'm manually setting --ignore-gpu-blacklist >>>> now. Not exactly the ideal user experience. >>>> >>>> On Tuesday, August 2, 2016 at 5:19:38 PM UTC-7, aaron...@gmail.com >>>> wrote: >>>>> >>>>> I've used WebGL on this computer for over a year with no problem, >>>>> until now. Now I'm in "the situation where Firefox and Safari can have >>>>> WebGL but Chrome doesn't" that you wanted to avoid... I've got that Intel >>>>> 3000 HD, on my 2011-era MBP. >>>> >>>> -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Ah, so this is why I no longer can load WebGL on Chrome? Seems a bit silly to blacklist it. I was experiencing no issues prior to this week, and suddenly I can't use WebGL anymore. I think it would be a better decision to allow its use, if it crashes on someone for the moment that's still a better outcome than not being able to use it at all. For most applications it's working perfectly fine for me. On Thursday, 16 June 2016 03:07:37 UTC+2, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
There should really be more awareness when making decisions like this that there are businesses that are depending on Chrome's WebGL support. We've generally recommended people use our product via Chrome, but will likely stop doing so if updates like these are going to be made so recklessly. On Wednesday, 15 June 2016 21:07:37 UTC-4, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 On Tue, Oct 4, 2016 at 8:50 AM, Tarek Sherif <tsherif@gmail.com> wrote: > There should really be more awareness when making decisions like this that > there are businesses that are depending on Chrome's WebGL support. We've > generally recommended people use our product via Chrome, but will likely > stop doing so if updates like these are going to be made so recklessly. > > > On Wednesday, 15 June 2016 21:07:37 UTC-4, ccam...@chromium.org wrote: >> >> Reviewers: Zhenyao Mo >> CL: https://codereview.chromium.org/2076443002/ >> >> Message: >> HD4000 is also on the chopping block, but this just does HD3000 for now. >> It >> holds the dubious honor of being responsible for 30% of our GPU crashes. >> >> Description: >> Blacklist HD3000 on Mac >> >> HD3000 is responsible for 30% of our GPU process crashes on Mac. >> >> BUG=592130 >> CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_g >> pu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_ >> rel;tryserver.chromium.win:win_optional_gpu_tests_rel >> >> Base URL: https://chromium.googlesource.com/chromium/src.git@master >> >> Affected files (+14, -1 lines): >> M gpu/config/software_rendering_list_json.cc >> >> >> Index: gpu/config/software_rendering_list_json.cc >> diff --git a/gpu/config/software_rendering_list_json.cc >> b/gpu/config/software_rendering_list_json.cc >> index 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 >> 100644 >> --- a/gpu/config/software_rendering_list_json.cc >> +++ b/gpu/config/software_rendering_list_json.cc >> @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = >> LONG_STRING_CONST( >> { >> "name": "software rendering list", >> // Please update the version number whenever you change this file. >> - "version": "11.1", >> + "version": "11.2", >> "entries": [ >> { >> "id": 1, >> @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( >> "features": [ >> "gpu_rasterization" >> ] >> + }, >> + { >> + "id": 112, >> + "description": "Intel HD 3000 driver crashes frequently on Mac", >> + "cr_bugs": [592130], >> + "os": { >> + "type": "macosx" >> + }, >> + "vendor_id": "0x8086", >> + "device_id": ["0x0116", "0x0126"], >> + "features": [ >> + "all" >> + ] >> } >> ] >> } >> >> >> -- > You received this message because you are subscribed to a topic in the > Google Groups "Chromium-reviews" group. > To unsubscribe from this topic, visit https://groups.google.com/a/ > chromium.org/d/topic/chromium-reviews/I6dL_9IAwiE/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > chromium-reviews+unsubscribe@chromium.org. > -- Peter Saal Phone: +1.303.578.9085 Skype: PLSAAL -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 on aggressive awful UX -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
This decision was not made recklessly or arbitrarily. This particular GPU was causing a disproportionately large percentage of Chrome's GPU process crashes on macOS and this had to be addressed. The data doesn't support doing this for the HD 4000 at this time so I've closed http://crbug.com/592130 to indicate that there are no current plans to blacklist that GPU. On 2016/10/04 14:50:17, tsherif_gmail.com wrote: > There should really be more awareness when making decisions like this that > there are businesses that are depending on Chrome's WebGL support. We've > generally recommended people use our product via Chrome, but will likely > stop doing so if updates like these are going to be made so recklessly. > > On Wednesday, 15 June 2016 21:07:37 UTC-4, mailto:ccam...@chromium.org wrote: > > > > Reviewers: Zhenyao Mo > > CL: https://codereview.chromium.org/2076443002/ > > > > Message: > > HD4000 is also on the chopping block, but this just does HD3000 for now. It > > holds the dubious honor of being responsible for 30% of our GPU crashes. > > > > Description: > > Blacklist HD3000 on Mac > > > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > > > BUG=592130 > > > > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > > > Affected files (+14, -1 lines): > > M gpu/config/software_rendering_list_json.cc > > > > > > Index: gpu/config/software_rendering_list_json.cc > > diff --git a/gpu/config/software_rendering_list_json.cc > > b/gpu/config/software_rendering_list_json.cc > > index > > > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > > > 100644 > > --- a/gpu/config/software_rendering_list_json.cc > > +++ b/gpu/config/software_rendering_list_json.cc > > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > > LONG_STRING_CONST( > > { > > "name": "software rendering list", > > // Please update the version number whenever you change this file. > > - "version": "11.1", > > + "version": "11.2", > > "entries": [ > > { > > "id": 1, > > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > > "features": [ > > "gpu_rasterization" > > ] > > + }, > > + { > > + "id": 112, > > + "description": "Intel HD 3000 driver crashes frequently on Mac", > > + "cr_bugs": [592130], > > + "os": { > > + "type": "macosx" > > + }, > > + "vendor_id": "0x8086", > > + "device_id": ["0x0116", "0x0126"], > > + "features": [ > > + "all" > > + ] > > } > > ] > > } > > > > > > > > -- > You received this message because you are subscribed to the Google Groups > "Chromium-reviews" group. > To unsubscribe from this group and stop receiving emails from it, send an email > to mailto:chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
What about a some sort of popup indicating that you've blocked something on the page due to X, if you wish to unblock it you have to manually continue. Way better than stopping it from working and giving everyone ugly cryptic errors. When the browser does this currently there is absolutely no way for the layperson to figure out what the problem is, and as a result web devs are the ones who cop the heat. How am I supposed to handle this on my side? I can't detect if a user is using this specific type of macbook pro and give them graceful degradation. By the way, expect a bunch more people in this thread as this type of macbook pro is still in use by a lot of developers. On Tue, Oct 4, 2016 at 8:29 PM, <kbr@chromium.org> wrote: > This decision was not made recklessly or arbitrarily. This particular GPU > was > causing a disproportionately large percentage of Chrome's GPU process > crashes on > macOS and this had to be addressed. The data doesn't support doing this > for the > HD 4000 at this time so I've closed http://crbug.com/592130 to indicate > that > there are no current plans to blacklist that GPU. > > On 2016/10/04 14:50:17, tsherif_gmail.com wrote: > > There should really be more awareness when making decisions like this > that > > there are businesses that are depending on Chrome's WebGL support. We've > > generally recommended people use our product via Chrome, but will likely > > stop doing so if updates like these are going to be made so recklessly. > > > > On Wednesday, 15 June 2016 21:07:37 UTC-4, mailto:ccam...@chromium.org > wrote: > > > > > > Reviewers: Zhenyao Mo > > > CL: https://codereview.chromium.org/2076443002/ > > > > > > Message: > > > HD4000 is also on the chopping block, but this just does HD3000 for > now. It > > > holds the dubious honor of being responsible for 30% of our GPU > crashes. > > > > > > Description: > > > Blacklist HD3000 on Mac > > > > > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > > > > > BUG=592130 > > > > > > > > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_ > gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_ > tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > > > > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > > > > > Affected files (+14, -1 lines): > > > M gpu/config/software_rendering_list_json.cc > > > > > > > > > Index: gpu/config/software_rendering_list_json.cc > > > diff --git a/gpu/config/software_rendering_list_json.cc > > > b/gpu/config/software_rendering_list_json.cc > > > index > > > > > > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a2 > 05ed50e6b5 > > > > > 100644 > > > --- a/gpu/config/software_rendering_list_json.cc > > > +++ b/gpu/config/software_rendering_list_json.cc > > > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > > > LONG_STRING_CONST( > > > { > > > "name": "software rendering list", > > > // Please update the version number whenever you change this file. > > > - "version": "11.1", > > > + "version": "11.2", > > > "entries": [ > > > { > > > "id": 1, > > > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > > > "features": [ > > > "gpu_rasterization" > > > ] > > > + }, > > > + { > > > + "id": 112, > > > + "description": "Intel HD 3000 driver crashes frequently on Mac", > > > + "cr_bugs": [592130], > > > + "os": { > > > + "type": "macosx" > > > + }, > > > + "vendor_id": "0x8086", > > > + "device_id": ["0x0116", "0x0126"], > > > + "features": [ > > > + "all" > > > + ] > > > } > > > ] > > > } > > > > > > > > > > > > > -- > > You received this message because you are subscribed to the Google Groups > > "Chromium-reviews" group. > > To unsubscribe from this group and stop receiving emails from it, send an > email > > to mailto:chromium-reviews+unsubscribe@chromium.org. > > > > https://codereview.chromium.org/2076443002/ > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
It may not seem arbitrary to you. I publish a very numerically & graphically demanding WebGL app for computational chemistry. I've never (NEVER) had WebGL crash on my 2011 MBP. One day, I update Chrome and suddenly my app won't initialize WebGL. That feels arbitrary to me -- but I'm not sure how you're going to notify me in any other way. But it sure sucks. And I can't imagine how my users who are suddenly blacklisted must have felt. I'm not denying that there are crashes and don't claim to understand the guts of WebGL implementations, but it seems that it ought to be possible to understand what WebGL features are causing the crashes. And that there are (ought to be?) ways for a developer to see if those features are supported on a particular GPU and work around them. I'm now using ignore-blacklist (with which I have to start Chrome from a shell, correct?) And my app continues to work fine (it has NEVER crashed WebGL, in case I forgot to mention that). And it works fine on FireFox's and Safari's and even iOS Safari's (sh*tty) WebGL implementations. Not arbitrary? Maybe not to you, but I gotta tell you I sure feels arbitrary to me and to my users. Thanks for your response. On Tue, Oct 4, 2016 at 1:29 PM, <kbr@chromium.org> wrote: > This decision was not made recklessly or arbitrarily. This particular GPU > was > causing a disproportionately large percentage of Chrome's GPU process > crashes on > macOS and this had to be addressed. The data doesn't support doing this > for the > HD 4000 at this time so I've closed http://crbug.com/592130 to indicate > that > there are no current plans to blacklist that GPU. > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
"This decision was not made recklessly or arbitrarily." https://youtu.be/pWdd6_ZxX8c On Tue, Oct 4, 2016 at 12:29 PM, <kbr@chromium.org> wrote: > This decision was not made recklessly or arbitrarily. This particular GPU > was > causing a disproportionately large percentage of Chrome's GPU process > crashes on > macOS and this had to be addressed. The data doesn't support doing this > for the > HD 4000 at this time so I've closed http://crbug.com/592130 to indicate > that > there are no current plans to blacklist that GPU. > > On 2016/10/04 14:50:17, tsherif_gmail.com wrote: > > There should really be more awareness when making decisions like this > that > > there are businesses that are depending on Chrome's WebGL support. We've > > generally recommended people use our product via Chrome, but will likely > > stop doing so if updates like these are going to be made so recklessly. > > > > On Wednesday, 15 June 2016 21:07:37 UTC-4, mailto:ccam...@chromium.org > wrote: > > > > > > Reviewers: Zhenyao Mo > > > CL: https://codereview.chromium.org/2076443002/ > > > > > > Message: > > > HD4000 is also on the chopping block, but this just does HD3000 for > now. It > > > holds the dubious honor of being responsible for 30% of our GPU > crashes. > > > > > > Description: > > > Blacklist HD3000 on Mac > > > > > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > > > > > BUG=592130 > > > > > > > > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_ > gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_ > tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > > > > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > > > > > Affected files (+14, -1 lines): > > > M gpu/config/software_rendering_list_json.cc > > > > > > > > > Index: gpu/config/software_rendering_list_json.cc > > > diff --git a/gpu/config/software_rendering_list_json.cc > > > b/gpu/config/software_rendering_list_json.cc > > > index > > > > > > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a2 > 05ed50e6b5 > > > > > 100644 > > > --- a/gpu/config/software_rendering_list_json.cc > > > +++ b/gpu/config/software_rendering_list_json.cc > > > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > > > LONG_STRING_CONST( > > > { > > > "name": "software rendering list", > > > // Please update the version number whenever you change this file. > > > - "version": "11.1", > > > + "version": "11.2", > > > "entries": [ > > > { > > > "id": 1, > > > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > > > "features": [ > > > "gpu_rasterization" > > > ] > > > + }, > > > + { > > > + "id": 112, > > > + "description": "Intel HD 3000 driver crashes frequently on Mac", > > > + "cr_bugs": [592130], > > > + "os": { > > > + "type": "macosx" > > > + }, > > > + "vendor_id": "0x8086", > > > + "device_id": ["0x0116", "0x0126"], > > > + "features": [ > > > + "all" > > > + ] > > > } > > > ] > > > } > > > > > > > > > > > > > -- > > You received this message because you are subscribed to the Google Groups > > "Chromium-reviews" group. > > To unsubscribe from this group and stop receiving emails from it, send an > email > > to mailto:chromium-reviews+unsubscribe@chromium.org. > > > > https://codereview.chromium.org/2076443002/ > -- Peter Saal Phone: +1.303.578.9085 Skype: PLSAAL -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Ken, is there some channel this was announced in that I'm not aware of? How are you communicating these decisions to people who depend on Chrome and WebGL? Tarek Sherif http://tareksherif.net/ https://www.biodigital.com/ On Tue, Oct 4, 2016 at 3:22 PM, Peter Saal <peter.saal@gmail.com> wrote: > "This decision was not made recklessly or arbitrarily." > https://youtu.be/pWdd6_ZxX8c > > On Tue, Oct 4, 2016 at 12:29 PM, <kbr@chromium.org> wrote: > >> This decision was not made recklessly or arbitrarily. This particular GPU >> was >> causing a disproportionately large percentage of Chrome's GPU process >> crashes on >> macOS and this had to be addressed. The data doesn't support doing this >> for the >> HD 4000 at this time so I've closed http://crbug.com/592130 to indicate >> that >> there are no current plans to blacklist that GPU. >> >> On 2016/10/04 14:50:17, tsherif_gmail.com wrote: >> > There should really be more awareness when making decisions like this >> that >> > there are businesses that are depending on Chrome's WebGL support. >> We've >> > generally recommended people use our product via Chrome, but will >> likely >> > stop doing so if updates like these are going to be made so recklessly. >> > >> > On Wednesday, 15 June 2016 21:07:37 UTC-4, mailto:ccam...@chromium.org >> wrote: >> > > >> > > Reviewers: Zhenyao Mo >> > > CL: https://codereview.chromium.org/2076443002/ >> > > >> > > Message: >> > > HD4000 is also on the chopping block, but this just does HD3000 for >> now. It >> > > holds the dubious honor of being responsible for 30% of our GPU >> crashes. >> > > >> > > Description: >> > > Blacklist HD3000 on Mac >> > > >> > > HD3000 is responsible for 30% of our GPU process crashes on Mac. >> > > >> > > BUG=592130 >> > > >> > > >> > >> CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_g >> pu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_ >> rel;tryserver.chromium.win:win_optional_gpu_tests_rel >> > > >> > > Base URL: https://chromium.googlesource.com/chromium/src.git@master >> > > >> > > Affected files (+14, -1 lines): >> > > M gpu/config/software_rendering_list_json.cc >> > > >> > > >> > > Index: gpu/config/software_rendering_list_json.cc >> > > diff --git a/gpu/config/software_rendering_list_json.cc >> > > b/gpu/config/software_rendering_list_json.cc >> > > index >> > > >> > >> 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a1 >> 5b43dbc1b1a205ed50e6b5 >> > >> > > 100644 >> > > --- a/gpu/config/software_rendering_list_json.cc >> > > +++ b/gpu/config/software_rendering_list_json.cc >> > > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = >> > > LONG_STRING_CONST( >> > > { >> > > "name": "software rendering list", >> > > // Please update the version number whenever you change this file. >> > > - "version": "11.1", >> > > + "version": "11.2", >> > > "entries": [ >> > > { >> > > "id": 1, >> > > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( >> > > "features": [ >> > > "gpu_rasterization" >> > > ] >> > > + }, >> > > + { >> > > + "id": 112, >> > > + "description": "Intel HD 3000 driver crashes frequently on Mac", >> > > + "cr_bugs": [592130], >> > > + "os": { >> > > + "type": "macosx" >> > > + }, >> > > + "vendor_id": "0x8086", >> > > + "device_id": ["0x0116", "0x0126"], >> > > + "features": [ >> > > + "all" >> > > + ] >> > > } >> > > ] >> > > } >> > > >> > > >> > > >> > >> > -- >> > You received this message because you are subscribed to the Google >> Groups >> > "Chromium-reviews" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> an >> email >> > to mailto:chromium-reviews+unsubscribe@chromium.org. >> >> >> >> https://codereview.chromium.org/2076443002/ >> > > > > -- > Peter Saal > Phone: +1.303.578.9085 > Skype: PLSAAL > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
On 2016/10/04 19:50:09, tsherif_gmail.com wrote: > Ken, is there some channel this was announced in that I'm not aware of? How > are you communicating these decisions to people who depend on Chrome and > WebGL? I'm afraid it was not announced publicly. Chrome users can always test the Beta, Dev and Canary channels to see what the planned upcoming behavior is and file bugs about it. We take our customers' feedback seriously and aim to address it. In this case it seems clear none of the affected users were testing the early access release channels. To be clear, this particular blacklist entry disables _all_ GPU acceleration features on the Intel HD 3000 on macOS. WebGL was not arbitrarily singled out. Chrome's crash reports were flooded with random crashes inside the graphics driver and there was no workaround that could be postulated from the Chrome team's standpoint aside from ceasing to utilize the GPU. One can use the command-line flag to ignore the GPU blacklist, though it should be understood that that is not a supported code path. This isn't the right forum on which to carry on this conversation so may I suggest either the webgl-dev-list Google group or public_webgl Khronos mailing list for follow-ups.
Message was sent while issue was closed.
> > This isn't the right forum on which to carry on this conversation so may I > suggest either the webgl-dev-list Google group or public_webgl Khronos > mailing > list for follow-ups. Neither of those seem like the correct forums either as this isn't a WebGL spec issue, nor a general WebGL development issue. It's more about communication between the Chrome team and their community. I'd suggest, though, that the webgl-dev-list would be a great place to announce these kinds of decisions to the community in the future. Tarek Sherif http://tareksherif.net/ https://www.biodigital.com/ On Tue, Oct 4, 2016 at 4:06 PM, <kbr@chromium.org> wrote: > On 2016/10/04 19:50:09, tsherif_gmail.com wrote: > > Ken, is there some channel this was announced in that I'm not aware of? > How > > are you communicating these decisions to people who depend on Chrome and > > WebGL? > > I'm afraid it was not announced publicly. Chrome users can always test the > Beta, > Dev and Canary channels to see what the planned upcoming behavior is and > file > bugs about it. We take our customers' feedback seriously and aim to > address it. > In this case it seems clear none of the affected users were testing the > early > access release channels. > > To be clear, this particular blacklist entry disables _all_ GPU > acceleration > features on the Intel HD 3000 on macOS. WebGL was not arbitrarily singled > out. > Chrome's crash reports were flooded with random crashes inside the graphics > driver and there was no workaround that could be postulated from the Chrome > team's standpoint aside from ceasing to utilize the GPU. One can use the > command-line flag to ignore the GPU blacklist, though it should be > understood > that that is not a supported code path. > > This isn't the right forum on which to carry on this conversation so may I > suggest either the webgl-dev-list Google group or public_webgl Khronos > mailing > list for follow-ups. > > > https://codereview.chromium.org/2076443002/ > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
So why is it that WebGL doesnt work at all even with the ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome anymore. Z. On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <tsherif@gmail.com> wrote: > This isn't the right forum on which to carry on this conversation so may I > > suggest either the webgl-dev-list Google group or public_webgl Khronos > mailing > list for follow-ups. > > > Neither of those seem like the correct forums either as this isn't a WebGL > spec issue, nor a general WebGL development issue. It's more about > communication between the Chrome team and their community. I'd suggest, > though, that the webgl-dev-list would be a great place to announce these > kinds of decisions to the community in the future. > > Tarek Sherif > http://tareksherif.net/ > https://www.biodigital.com/ > > > On Tue, Oct 4, 2016 at 4:06 PM, <kbr@chromium.org> wrote: > > On 2016/10/04 19:50:09, tsherif_gmail.com wrote: > > Ken, is there some channel this was announced in that I'm not aware of? > How > > are you communicating these decisions to people who depend on Chrome and > > WebGL? > > I'm afraid it was not announced publicly. Chrome users can always test the > Beta, > Dev and Canary channels to see what the planned upcoming behavior is and > file > bugs about it. We take our customers' feedback seriously and aim to > address it. > In this case it seems clear none of the affected users were testing the > early > access release channels. > > To be clear, this particular blacklist entry disables _all_ GPU > acceleration > features on the Intel HD 3000 on macOS. WebGL was not arbitrarily singled > out. > Chrome's crash reports were flooded with random crashes inside the graphics > driver and there was no workaround that could be postulated from the Chrome > team's standpoint aside from ceasing to utilize the GPU. One can use the > command-line flag to ignore the GPU blacklist, though it should be > understood > that that is not a supported code path. > > This isn't the right forum on which to carry on this conversation so may I > suggest either the webgl-dev-list Google group or public_webgl Khronos > mailing > list for follow-ups. > > > https://codereview.chromium.org/2076443002/ > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Please provide the about:gpu content when you run chrome with --ignore-gpu-blacklist. On Wed, Oct 5, 2016 at 4:36 AM, Páll Zoltán <pallzoltan@gmail.com> wrote: > So why is it that WebGL doesnt work at all even with the > ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome > anymore. > > Z. > > On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <tsherif@gmail.com> wrote: >>> >>> This isn't the right forum on which to carry on this conversation so may >>> I >>> >>> suggest either the webgl-dev-list Google group or public_webgl Khronos >>> mailing >>> list for follow-ups. >> >> >> Neither of those seem like the correct forums either as this isn't a WebGL >> spec issue, nor a general WebGL development issue. It's more about >> communication between the Chrome team and their community. I'd suggest, >> though, that the webgl-dev-list would be a great place to announce these >> kinds of decisions to the community in the future. >> >> Tarek Sherif >> http://tareksherif.net/ >> https://www.biodigital.com/ >> >> >> On Tue, Oct 4, 2016 at 4:06 PM, <kbr@chromium.org> wrote: >>> >>> On 2016/10/04 19:50:09, tsherif_gmail.com wrote: >>> > Ken, is there some channel this was announced in that I'm not aware of? >>> > How >>> > are you communicating these decisions to people who depend on Chrome >>> > and >>> > WebGL? >>> >>> I'm afraid it was not announced publicly. Chrome users can always test >>> the Beta, >>> Dev and Canary channels to see what the planned upcoming behavior is and >>> file >>> bugs about it. We take our customers' feedback seriously and aim to >>> address it. >>> In this case it seems clear none of the affected users were testing the >>> early >>> access release channels. >>> >>> To be clear, this particular blacklist entry disables _all_ GPU >>> acceleration >>> features on the Intel HD 3000 on macOS. WebGL was not arbitrarily singled >>> out. >>> Chrome's crash reports were flooded with random crashes inside the >>> graphics >>> driver and there was no workaround that could be postulated from the >>> Chrome >>> team's standpoint aside from ceasing to utilize the GPU. One can use the >>> command-line flag to ignore the GPU blacklist, though it should be >>> understood >>> that that is not a supported code path. >>> >>> This isn't the right forum on which to carry on this conversation so may >>> I >>> suggest either the webgl-dev-list Google group or public_webgl Khronos >>> mailing >>> list for follow-ups. >>> >>> >>> https://codereview.chromium.org/2076443002/ >> >> > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Here it is, Zhenyao: http://pastebin.com/VAiA8Cbt <http://pastebin.com/HbRdCb3f> with some logs included. Google Maps always crashes WebGL. Z. On Wed, Oct 5, 2016 at 8:09 PM Zhenyao Mo <zmo@chromium.org> wrote: > Please provide the about:gpu content when you run chrome with > --ignore-gpu-blacklist. > > On Wed, Oct 5, 2016 at 4:36 AM, Páll Zoltán <pallzoltan@gmail.com> wrote: > > So why is it that WebGL doesnt work at all even with the > > ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome > > anymore. > > > > Z. > > > > On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <tsherif@gmail.com> wrote: > >>> > >>> This isn't the right forum on which to carry on this conversation so > may > >>> I > >>> > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > >>> mailing > >>> list for follow-ups. > >> > >> > >> Neither of those seem like the correct forums either as this isn't a > WebGL > >> spec issue, nor a general WebGL development issue. It's more about > >> communication between the Chrome team and their community. I'd suggest, > >> though, that the webgl-dev-list would be a great place to announce these > >> kinds of decisions to the community in the future. > >> > >> Tarek Sherif > >> http://tareksherif.net/ > >> https://www.biodigital.com/ > >> > >> > >> On Tue, Oct 4, 2016 at 4:06 PM, <kbr@chromium.org> wrote: > >>> > >>> On 2016/10/04 19:50:09, tsherif_gmail.com wrote: > >>> > Ken, is there some channel this was announced in that I'm not aware > of? > >>> > How > >>> > are you communicating these decisions to people who depend on Chrome > >>> > and > >>> > WebGL? > >>> > >>> I'm afraid it was not announced publicly. Chrome users can always test > >>> the Beta, > >>> Dev and Canary channels to see what the planned upcoming behavior is > and > >>> file > >>> bugs about it. We take our customers' feedback seriously and aim to > >>> address it. > >>> In this case it seems clear none of the affected users were testing the > >>> early > >>> access release channels. > >>> > >>> To be clear, this particular blacklist entry disables _all_ GPU > >>> acceleration > >>> features on the Intel HD 3000 on macOS. WebGL was not arbitrarily > singled > >>> out. > >>> Chrome's crash reports were flooded with random crashes inside the > >>> graphics > >>> driver and there was no workaround that could be postulated from the > >>> Chrome > >>> team's standpoint aside from ceasing to utilize the GPU. One can use > the > >>> command-line flag to ignore the GPU blacklist, though it should be > >>> understood > >>> that that is not a supported code path. > >>> > >>> This isn't the right forum on which to carry on this conversation so > may > >>> I > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > >>> mailing > >>> list for follow-ups. > >>> > >>> > >>> https://codereview.chromium.org/2076443002/ > >> > >> > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Might be worthwhile investigating why this is the case. Google maps webgl has never crashed my browser. What potential differences are there? On 6 Oct. 2016 15:56, "Páll Zoltán" <pallzoltan@gmail.com> wrote: Here it is, Zhenyao: http://pastebin.com/VAiA8Cbt <http://pastebin.com/HbRdCb3f> with some logs included. Google Maps always crashes WebGL. Z. On Wed, Oct 5, 2016 at 8:09 PM Zhenyao Mo <zmo@chromium.org> wrote: > Please provide the about:gpu content when you run chrome with > --ignore-gpu-blacklist. > > On Wed, Oct 5, 2016 at 4:36 AM, Páll Zoltán <pallzoltan@gmail.com> wrote: > > So why is it that WebGL doesnt work at all even with the > > ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome > > anymore. > > > > Z. > > > > On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <tsherif@gmail.com> wrote: > >>> > >>> This isn't the right forum on which to carry on this conversation so > may > >>> I > >>> > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > >>> mailing > >>> list for follow-ups. > >> > >> > >> Neither of those seem like the correct forums either as this isn't a > WebGL > >> spec issue, nor a general WebGL development issue. It's more about > >> communication between the Chrome team and their community. I'd suggest, > >> though, that the webgl-dev-list would be a great place to announce these > >> kinds of decisions to the community in the future. > >> > >> Tarek Sherif > >> http://tareksherif.net/ > >> https://www.biodigital.com/ > >> > >> > >> On Tue, Oct 4, 2016 at 4:06 PM, <kbr@chromium.org> wrote: > >>> > >>> On 2016/10/04 19:50:09, tsherif_gmail.com wrote: > >>> > Ken, is there some channel this was announced in that I'm not aware > of? > >>> > How > >>> > are you communicating these decisions to people who depend on Chrome > >>> > and > >>> > WebGL? > >>> > >>> I'm afraid it was not announced publicly. Chrome users can always test > >>> the Beta, > >>> Dev and Canary channels to see what the planned upcoming behavior is > and > >>> file > >>> bugs about it. We take our customers' feedback seriously and aim to > >>> address it. > >>> In this case it seems clear none of the affected users were testing the > >>> early > >>> access release channels. > >>> > >>> To be clear, this particular blacklist entry disables _all_ GPU > >>> acceleration > >>> features on the Intel HD 3000 on macOS. WebGL was not arbitrarily > singled > >>> out. > >>> Chrome's crash reports were flooded with random crashes inside the > >>> graphics > >>> driver and there was no workaround that could be postulated from the > >>> Chrome > >>> team's standpoint aside from ceasing to utilize the GPU. One can use > the > >>> command-line flag to ignore the GPU blacklist, though it should be > >>> understood > >>> that that is not a supported code path. > >>> > >>> This isn't the right forum on which to carry on this conversation so > may > >>> I > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > >>> mailing > >>> list for follow-ups. > >>> > >>> > >>> https://codereview.chromium.org/2076443002/ > >> > >> > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
On 2016/10/06 13:56:11, pallzoltan_gmail.com wrote: > Here it is, Zhenyao: http://pastebin.com/VAiA8Cbt > <http://pastebin.com/HbRdCb3f> with some logs included. > Google Maps always crashes WebGL. > > Z. > > On Wed, Oct 5, 2016 at 8:09 PM Zhenyao Mo <mailto:zmo@chromium.org> wrote: > > > Please provide the about:gpu content when you run chrome with > > --ignore-gpu-blacklist. > > > > On Wed, Oct 5, 2016 at 4:36 AM, Páll Zoltán <mailto:pallzoltan@gmail.com> wrote: > > > So why is it that WebGL doesnt work at all even with the > > > ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome > > > anymore. > > > > > > Z. > > > > > > On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <mailto:tsherif@gmail.com> wrote: > > >>> > > >>> This isn't the right forum on which to carry on this conversation so > > may > > >>> I > > >>> > > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > > >>> mailing > > >>> list for follow-ups. > > >> > > >> > > >> Neither of those seem like the correct forums either as this isn't a > > WebGL > > >> spec issue, nor a general WebGL development issue. It's more about > > >> communication between the Chrome team and their community. I'd suggest, > > >> though, that the webgl-dev-list would be a great place to announce these > > >> kinds of decisions to the community in the future. > > >> > > >> Tarek Sherif > > >> http://tareksherif.net/ > > >> https://www.biodigital.com/ > > >> > > >> > > >> On Tue, Oct 4, 2016 at 4:06 PM, <mailto:kbr@chromium.org> wrote: > > >>> > > >>> On 2016/10/04 19:50:09, http://tsherif_gmail.com wrote: > > >>> > Ken, is there some channel this was announced in that I'm not aware > > of? > > >>> > How > > >>> > are you communicating these decisions to people who depend on Chrome > > >>> > and > > >>> > WebGL? > > >>> > > >>> I'm afraid it was not announced publicly. Chrome users can always test > > >>> the Beta, > > >>> Dev and Canary channels to see what the planned upcoming behavior is > > and > > >>> file > > >>> bugs about it. We take our customers' feedback seriously and aim to > > >>> address it. > > >>> In this case it seems clear none of the affected users were testing the > > >>> early > > >>> access release channels. > > >>> > > >>> To be clear, this particular blacklist entry disables _all_ GPU > > >>> acceleration > > >>> features on the Intel HD 3000 on macOS. WebGL was not arbitrarily > > singled > > >>> out. > > >>> Chrome's crash reports were flooded with random crashes inside the > > >>> graphics > > >>> driver and there was no workaround that could be postulated from the > > >>> Chrome > > >>> team's standpoint aside from ceasing to utilize the GPU. One can use > > the > > >>> command-line flag to ignore the GPU blacklist, though it should be > > >>> understood > > >>> that that is not a supported code path. > > >>> > > >>> This isn't the right forum on which to carry on this conversation so > > may > > >>> I > > >>> suggest either the webgl-dev-list Google group or public_webgl Khronos > > >>> mailing > > >>> list for follow-ups. > > >>> > > >>> > > >>> https://codereview.chromium.org/2076443002/ > > >> > > >> > > > > > > > -- > You received this message because you are subscribed to the Google Groups > "Chromium-reviews" group. > To unsubscribe from this group and stop receiving emails from it, send an email > to mailto:chromium-reviews+unsubscribe@chromium.org. It seems you are using core profile (because you specified --enable-unsafe-es3-apis) rather than the general compatibility profile GL. Can you try without --enable-unsafe-es3-apis? We are not going to support WebGL2 on this hardware anyway even without high crash rate because it's only GL 3.3. The lowest GL we are able to run WebGL2 will be 4.1 (even that is a stretch because we have to work around a bunch of missing stuff or inconsistency with ES3 that's not there in later GL versions).
Message was sent while issue was closed.
That helped, I have WebGL back. Thank you so much, zmo. I'm absolutely happy with this solution. Also, for the record, I've never had problems with WebGL on this hardware. Thank you once more, Z. On Thu, Oct 6, 2016 at 8:57 PM <zmo@chromium.org> wrote: > On 2016/10/06 13:56:11, pallzoltan_gmail.com wrote: > > Here it is, Zhenyao: http://pastebin.com/VAiA8Cbt > > <http://pastebin.com/HbRdCb3f> with some logs included. > > > Google Maps always crashes WebGL. > > > > Z. > > > > On Wed, Oct 5, 2016 at 8:09 PM Zhenyao Mo <mailto:zmo@chromium.org> > wrote: > > > > > Please provide the about:gpu content when you run chrome with > > > --ignore-gpu-blacklist. > > > > > > On Wed, Oct 5, 2016 at 4:36 AM, Páll Zoltán <mailto: > pallzoltan@gmail.com> > > wrote: > > > > So why is it that WebGL doesnt work at all even with the > > > > ignore-gpu-blacklist flag? Whatever I do, WebGL won't work in Chrome > > > > anymore. > > > > > > > > Z. > > > > > > > > On Tue, Oct 4, 2016 at 11:49 PM Tarek Sherif <mailto: > tsherif@gmail.com> > > wrote: > > > >>> > > > >>> This isn't the right forum on which to carry on this conversation > so > > > may > > > >>> I > > > >>> > > > >>> suggest either the webgl-dev-list Google group or public_webgl > Khronos > > > >>> mailing > > > >>> list for follow-ups. > > > >> > > > >> > > > >> Neither of those seem like the correct forums either as this isn't a > > > WebGL > > > >> spec issue, nor a general WebGL development issue. It's more about > > > >> communication between the Chrome team and their community. I'd > suggest, > > > >> though, that the webgl-dev-list would be a great place to announce > these > > > >> kinds of decisions to the community in the future. > > > >> > > > >> Tarek Sherif > > > >> http://tareksherif.net/ > > > >> https://www.biodigital.com/ > > > >> > > > >> > > > >> On Tue, Oct 4, 2016 at 4:06 PM, <mailto:kbr@chromium.org> wrote: > > > > >>> > > > >>> On 2016/10/04 19:50:09, http://tsherif_gmail.com wrote: > > > >>> > Ken, is there some channel this was announced in that I'm not > aware > > > of? > > > >>> > How > > > >>> > are you communicating these decisions to people who depend on > Chrome > > > >>> > and > > > >>> > WebGL? > > > >>> > > > >>> I'm afraid it was not announced publicly. Chrome users can always > test > > > >>> the Beta, > > > >>> Dev and Canary channels to see what the planned upcoming behavior > is > > > and > > > >>> file > > > >>> bugs about it. We take our customers' feedback seriously and aim to > > > >>> address it. > > > >>> In this case it seems clear none of the affected users were > testing the > > > >>> early > > > >>> access release channels. > > > >>> > > > >>> To be clear, this particular blacklist entry disables _all_ GPU > > > >>> acceleration > > > >>> features on the Intel HD 3000 on macOS. WebGL was not arbitrarily > > > singled > > > >>> out. > > > >>> Chrome's crash reports were flooded with random crashes inside the > > > >>> graphics > > > >>> driver and there was no workaround that could be postulated from > the > > > >>> Chrome > > > >>> team's standpoint aside from ceasing to utilize the GPU. One can > use > > > the > > > >>> command-line flag to ignore the GPU blacklist, though it should be > > > >>> understood > > > >>> that that is not a supported code path. > > > >>> > > > >>> This isn't the right forum on which to carry on this conversation > so > > > may > > > >>> I > > > >>> suggest either the webgl-dev-list Google group or public_webgl > Khronos > > > >>> mailing > > > >>> list for follow-ups. > > > >>> > > > >>> > > > >>> https://codereview.chromium.org/2076443002/ > > > >> > > > >> > > > > > > > > > > > -- > > You received this message because you are subscribed to the Google Groups > > "Chromium-reviews" group. > > To unsubscribe from this group and stop receiving emails from it, send an > email > > to mailto:chromium-reviews+unsubscribe@chromium.org. > > It seems you are using core profile (because you specified > --enable-unsafe-es3-apis) rather than the general compatibility profile GL. > > Can you try without --enable-unsafe-es3-apis? We are not going to support > WebGL2 on this hardware anyway even without high crash rate because it's > only GL > 3.3. The lowest GL we are able to run WebGL2 will be 4.1 (even that is a > stretch > because we have to work around a bunch of missing stuff or inconsistency > with > ES3 that's not there in later GL versions). > > https://codereview.chromium.org/2076443002/ > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Same here. Never had issues with my MBP2011 running webgl content, but no... Chrome devs had to ruin the day without a hint of warning!! Totally irresponsible decision and just looks like laziness on their behalf. Τη Πέμπτη, 16 Ιουνίου 2016 - 4:07:37 π.μ. UTC+3, ο χρήστης ccam...@chromium.org έγραψε: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 for being sad about this. I totally understand the reasons behind wanting to disable it for an overall less crashy user experience, but we've made several technical decisions based on decent webGL support being wide spread, and now it seems like we may have to reconsider because of this. Is there some way of letting a user enable/disable webGL for each tab and/or rendering instance, just like how flash works? At least then there would be a nicer UI for opting into a potentially crashy experience rather than having to set it via options. On Wednesday, June 15, 2016 at 6:07:37 PM UTC-7, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
I have the same problem on Macbook Air 2011 with HD3000. It was running fine until yesterday when the HD300 was blacklisted. I don't understand why Safari works but not Chrome? On Wednesday, June 15, 2016 at 9:07:37 PM UTC-4, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Can there be some detail behind this, please? I don't understand how such a drastic change can be made without clear reasoning or any data for us to go on. Tanking a majority/70% of a user base (probably higher due to infrequency?) without warning or information should not be handled this way. It seems like a significant blow to WebGL as a platform, let alone canvas etc... Accountability is vital here, and this change feels heavy handed to say the least. - On Wednesday, November 23, 2016 at 8:49:46 PM UTC-3:30, tir...@gmail.com wrote: > > > I have the same problem on Macbook Air 2011 with HD3000. It was running > fine until yesterday when the HD300 was blacklisted. I don't understand > why Safari works but not Chrome? > > > On Wednesday, June 15, 2016 at 9:07:37 PM UTC-4, ccam...@chromium.org > wrote: >> >> Reviewers: Zhenyao Mo >> CL: https://codereview.chromium.org/2076443002/ >> >> Message: >> HD4000 is also on the chopping block, but this just does HD3000 for now. >> It >> holds the dubious honor of being responsible for 30% of our GPU crashes. >> >> Description: >> Blacklist HD3000 on Mac >> >> HD3000 is responsible for 30% of our GPU process crashes on Mac. >> >> BUG=592130 >> >> CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel >> >> Base URL: https://chromium.googlesource.com/chromium/src.git@master >> >> Affected files (+14, -1 lines): >> M gpu/config/software_rendering_list_json.cc >> >> >> Index: gpu/config/software_rendering_list_json.cc >> diff --git a/gpu/config/software_rendering_list_json.cc >> b/gpu/config/software_rendering_list_json.cc >> index >> 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 >> 100644 >> --- a/gpu/config/software_rendering_list_json.cc >> +++ b/gpu/config/software_rendering_list_json.cc >> @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = >> LONG_STRING_CONST( >> { >> "name": "software rendering list", >> // Please update the version number whenever you change this file. >> - "version": "11.1", >> + "version": "11.2", >> "entries": [ >> { >> "id": 1, >> @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( >> "features": [ >> "gpu_rasterization" >> ] >> + }, >> + { >> + "id": 112, >> + "description": "Intel HD 3000 driver crashes frequently on Mac", >> + "cr_bugs": [592130], >> + "os": { >> + "type": "macosx" >> + }, >> + "vendor_id": "0x8086", >> + "device_id": ["0x0116", "0x0126"], >> + "features": [ >> + "all" >> + ] >> } >> ] >> } >> >> >> -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
I hope this issue gets fixed. Now I can't code Three.js stuff and use Chrome? WebGL works fine on Firefox and Safari so please optimize the code on Chrome. WebGL never crashed for me on HD3000. I thought Chrome was the better browser. On Wednesday, June 15, 2016 at 6:07:37 PM UTC-7, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
Hi guys, I'm back again because few days ago WebGL started crashing again. I've been using Chrome with *ignore-gpu-blacklist* and everything worked fine. Due to some update, Google Maps won't work with WebGL anymore, it just crashes and falls back to the simple version. Any idea what's going on? Here's my *chrome://gpu* output: http://pastebin.com/2XqqwfAk Cheers, Z. On Tue, Dec 20, 2016 at 7:20 AM <dub909@gmail.com> wrote: > I hope this issue gets fixed. Now I can't code Three.js stuff and use > Chrome? WebGL works fine on Firefox and Safari so please optimize the code > on Chrome. WebGL never crashed for me on HD3000. I thought Chrome was the > better browser. > > > On Wednesday, June 15, 2016 at 6:07:37 PM UTC-7, ccam...@chromium.org > wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- > You received this message because you are subscribed to a topic in the > Google Groups "Chromium-reviews" group. > To unsubscribe from this topic, visit > https://groups.google.com/a/chromium.org/d/topic/chromium-reviews/I6dL_9IAwiE... > . > To unsubscribe from this group and all its topics, send an email to > chromium-reviews+unsubscribe@chromium.org. > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 too aggressive. WebGL has worked fine for years (late 2011 MBP). Now without explicit explanation Chrome has no WebGL support. Computers are not becoming obsolete as fast as they used to. On Thursday, June 16, 2016 at 3:07:37 AM UTC+2, ccam...@chromium.org wrote: > > Reviewers: Zhenyao Mo > CL: https://codereview.chromium.org/2076443002/ > > Message: > HD4000 is also on the chopping block, but this just does HD3000 for now. It > holds the dubious honor of being responsible for 30% of our GPU crashes. > > Description: > Blacklist HD3000 on Mac > > HD3000 is responsible for 30% of our GPU process crashes on Mac. > > BUG=592130 > > CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel > > Base URL: https://chromium.googlesource.com/chromium/src.git@master > > Affected files (+14, -1 lines): > M gpu/config/software_rendering_list_json.cc > > > Index: gpu/config/software_rendering_list_json.cc > diff --git a/gpu/config/software_rendering_list_json.cc > b/gpu/config/software_rendering_list_json.cc > index > 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 > 100644 > --- a/gpu/config/software_rendering_list_json.cc > +++ b/gpu/config/software_rendering_list_json.cc > @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = > LONG_STRING_CONST( > { > "name": "software rendering list", > // Please update the version number whenever you change this file. > - "version": "11.1", > + "version": "11.2", > "entries": [ > { > "id": 1, > @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( > "features": [ > "gpu_rasterization" > ] > + }, > + { > + "id": 112, > + "description": "Intel HD 3000 driver crashes frequently on Mac", > + "cr_bugs": [592130], > + "os": { > + "type": "macosx" > + }, > + "vendor_id": "0x8086", > + "device_id": ["0x0116", "0x0126"], > + "features": [ > + "all" > + ] > } > ] > } > > > -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org.
Message was sent while issue was closed.
+1 too aggressive. I had google map 3d working perfectly on MBP late 2011, but cant use it anymore after chrome update. Anyone knows what is the latest chrome version that doesnt black liste Intel HD 3000 grapic card? Le dimanche 15 janvier 2017 12:13:02 UTC+1, kte...@gmail.com a écrit : > > +1 too aggressive. WebGL has worked fine for years (late 2011 MBP). Now > without explicit explanation Chrome has no WebGL support. Computers are not > becoming obsolete as fast as they used to. > > On Thursday, June 16, 2016 at 3:07:37 AM UTC+2, ccam...@chromium.org > wrote: >> >> Reviewers: Zhenyao Mo >> CL: https://codereview.chromium.org/2076443002/ >> >> Message: >> HD4000 is also on the chopping block, but this just does HD3000 for now. >> It >> holds the dubious honor of being responsible for 30% of our GPU crashes. >> >> Description: >> Blacklist HD3000 on Mac >> >> HD3000 is responsible for 30% of our GPU process crashes on Mac. >> >> BUG=592130 >> >> CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel >> >> Base URL: https://chromium.googlesource.com/chromium/src.git@master >> >> Affected files (+14, -1 lines): >> M gpu/config/software_rendering_list_json.cc >> >> >> Index: gpu/config/software_rendering_list_json.cc >> diff --git a/gpu/config/software_rendering_list_json.cc >> b/gpu/config/software_rendering_list_json.cc >> index >> 1d5c2b161eed48188e140075fbbcb38c6b6cc105..ef5bad6b90f18fe5a15b43dbc1b1a205ed50e6b5 >> 100644 >> --- a/gpu/config/software_rendering_list_json.cc >> +++ b/gpu/config/software_rendering_list_json.cc >> @@ -18,7 +18,7 @@ const char kSoftwareRenderingListJson[] = >> LONG_STRING_CONST( >> { >> "name": "software rendering list", >> // Please update the version number whenever you change this file. >> - "version": "11.1", >> + "version": "11.2", >> "entries": [ >> { >> "id": 1, >> @@ -1075,6 +1075,19 @@ LONG_STRING_CONST( >> "features": [ >> "gpu_rasterization" >> ] >> + }, >> + { >> + "id": 112, >> + "description": "Intel HD 3000 driver crashes frequently on Mac", >> + "cr_bugs": [592130], >> + "os": { >> + "type": "macosx" >> + }, >> + "vendor_id": "0x8086", >> + "device_id": ["0x0116", "0x0126"], >> + "features": [ >> + "all" >> + ] >> } >> ] >> } >> >> >> -- You received this message because you are subscribed to the Google Groups "Chromium-reviews" group. To unsubscribe from this group and stop receiving emails from it, send an email to chromium-reviews+unsubscribe@chromium.org. |