View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0005815 | Spring engine | Gfx driver issues | public | 2017-10-21 09:56 | 2019-04-11 00:42 |
| Reporter | Sortale | Assigned To | Kloot | ||
| Priority | normal | Severity | major | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Product Version | 104.0 +git | ||||
| Fixed in Version | 104.0 +git | ||||
| Summary | 0005815: 104 engine render map with repeated square of texture on AMD Gfx card | ||||
| Description | same spring configure file on both 103 and 104 engine 103 no problem 104 bad map rendering with repeated square of texture. Manually enable compatibility in chobby fix this issue but break the nanothal/lazer rendering | ||||
| Tags | No tags attached. | ||||
| Attached Files | |||||
| Checked infolog.txt for Errors | |||||
|
|
"same spring configure file on both 103 and 104 engine" attach infologs, 99% chance the UsePBO setting will differ. |
|
|
infolog 104.0.1-49-g70f081f https://pastebin.com/8mWTzq0E infolog 103.0.1-540-g9fb8b24 https://pastebin.com/t1unR9r1 I literally copied the same springsettings.cfg file to both folder |
|
|
apparently we have a 1%er here |
|
|
then he'd better feel like performing a bisection because the changes to SMFGroundTextures between g9fb8b24 and g70f081f don't explain this. |
|
|
similar events have been reported by other AMD user. I do not believe this is rare. bisection? bisection?!!! I just cleaned my engine folder less than a week ago. you monster! joke aside if you are seriously looking into it, I'll try doing bisection. I was informed that this is a long-standing issue with known config fix and I was convinced that this is more of a config assigning/reading issue. But I have 0 experience in development so feel free to disregard everything I say. |
|
|
I locally have ~all windows versions between 98.0.1-226-gd9636d1 and latest. (only slightly above 2GB) in a git repo. I'll try to make a smaller repo containing only the relevant versions and upload it so you can download and bisect. I don't have maintenance branch in the repo, so please report what happens in 104.0. |
|
|
These engine versions are on the ZK server: https://pastebin.com/QV796pTu You can make the wrapper automatically download and run of of them by running the wrapper with the name of the engine you want to run (for example, chobby.exe 103.0.1-540-g9fb8b24). |
|
|
Bug is not present in 104.0.1-7-gf03115b, next version, that is available to me, 104.0.1-49-g70f081f, has this bug. Therefore it was introduced between 7 and 49. |
|
|
Ok, I bisected further: 1) 104.0.1-7-gf03115b - doesn't have bug 2) 104.0.1-12-geb35cc6 - has bug |
|
|
@lhog what OS are you testing on? |
|
|
Win10 x64 |
|
|
https://github.com/spring/spring/commit/eb35cc6 https://github.com/spring/spring/commit/9ca7fd3 https://github.com/spring/spring/commit/75b7e4a https://github.com/spring/spring/commit/a6a8be4 https://github.com/spring/spring/commit/e6aa016 https://github.com/spring/spring/commit/f03115b nothing at all related to graphics code so this is a giant WTF and I'm inclined to believe you have a false positive. |
|
|
glew didn't change between those two releases so I'm truly baffled |
|
|
Actually my test was performed in less than clean environment. I had 104.0.1-7-gf03115b downloaded prior to the tests. Despite it was placed in Win64 folder it was in fact Win32 version of engine and it indeed doesn't have the bug. 104.0.1-7-gf03115b/Win64 is bugged. |
|
|
the win64 version _does_ have the new glew version. Can you check only win32 versions? |
|
|
win64 buildbot Spring has various other issues (0005689 etc), it's not a good testing platform. on the other hand, since this bug dates back much further than 103.0.1-540-g9fb8b24 on win32, whatever commit introduced it for Sortale might at least provide a hint as to the real cause. |
|
|
On Win32 I see no map render issue up to and including 104.0.1-48-g1739d57. I can't launch 104.0.1-53-ge4def60 and 104.0.1-55-gf18b3f4 due to "GLEW_ARB_buffer_storage not supported". Note that Win64 counterparts do launch without GLEW error, but have this map rendering glitch. |
|
|
uhm GoogleFrog method of acquiring engine not working Google search show only this https://springrts.com/wiki/Download https://springrts.com/dl/buildbot/default/develop/ It should be noted that the same problem was found in this engine in a previous test https://springrts.com/dl/buildbot/default/develop/103.0.1-1169-gd576c41/ |
|
|
103.0.1-648-g5775975 do not have it. Info log below https://pastebin.com/ZKwNht2t |
|
|
From lhog's results I'm pretty sure it's the glew thing. I'll set up an exact test, you don't really have to go through all engines. |
|
|
103.0.1-1144-gcd36288_win64 have the bug, info log below https://pastebin.com/fPTNfcPP |
|
|
Sortale you can only download these versions with the wrapper as the rest are not hosted: https://pastebin.com/QV796pTu |
|
|
@lhog & @Sortale please test https://springrts.com/dl/buildbot/default/develop/104.0.1-285-g9252e85/win32/ https://springrts.com/dl/buildbot/default/develop/104.0.1-287-gce39a84/win32/ and if you have time, also https://springrts.com/dl/buildbot/default/develop/104.0.1-285-g9252e85/win64/ https://springrts.com/dl/buildbot/default/develop/104.0.1-287-gce39a84/win64/ thanks! |
|
|
@Google_Frog I used the list, I am not that silly [ok maybe I am usually that silly and your concern is justified but still] @hokomoko can't run any of those error message: OpenGL extension "GLEW_EXT_texture_compression_dxt1" not supported, aborting infolog https://pastebin.com/PJuAJEZn |
|
|
sorry, try this when it's up (should be in a minute or so) https://springrts.com/dl/buildbot/default/develop/104.0.1-288-g624cb6b/win32/ |
|
|
or https://springrts.com/dl/buildbot/default/develop/104.0.1-289-g62f4019/win32/ |
|
|
Not good (@289/win32) |
|
|
that is expected, map rendering in develop uses GL4 now. either update your win32 drivers or wait for a new win64 maintenance build. |
|
|
EDIT: nevermind, @lhog screenshot shows a different problem... weird. |
|
|
lhog's issue with 289/win32 (screen00029.jpg) is not a PBO failure... |
|
|
Yes, I only saw he added a screenshot after posting :) |
|
|
104.0.1-289-g62f4019 fail horribly I can start the game but here's the graphic also the loading screen is suddenly rendered in a tiny square lower left corner https://imgur.com/a/hkyWK same with 104.0.1-289-g62f4019 |
|
|
edit your springsettings.cfg and add these lines: GLContextMajorVersion=4 GLContextMinorVersion=5 if it still fails, then also add DebugGL=1 and post the infolog. |
|
|
Just have tested two options: 1. GLContextMajorVersion=4 GLContextMinorVersion=1 Doesn't change a thing. Still same visuals as before (similar to @Sortale https://imgur.com/a/hkyWK) 2. GLContextMajorVersion=4 GLContextMinorVersion=5 Fails right from the start: "GLEW version outdated, aborting" I have Radeon 5870 GPU and latest drivers provided by AMD. |
|
|
1) *always* attach an infolog. 2) the latest Radeon HD5xxx Windows driver from https://support.amd.com/en-us/download goes up to GL4.4 (so choose GLContextMinorVersion=4) which should include support for GLEW_ARB_buffer_storage required by develop since g6104061, if it does not you're out of luck. 3) as mentioned in https://springrts.com/mantis/view.php?id=5815#c18582 maintenance builds still use an older GLEW, hence the "version outdated" message. if you get this on develop 289-g62f4019 then the mingwlibs update must have failed in some way. |
|
|
Configuration with GLContextMajorVersion=4, GLContextMinorVersion=4 does start, but the map is same empty as before. Infolog attached. |
|
|
"Validation failed! - Different sampler types for same sample texture unit in fragment shader." ATI program validation logic seems to have been changed or broken, this is the GL4 incarnation of 0002624. |
|
|
(@lhog and @Sortale) retry with 295-g1ac4c81. |
|
|
Tried 295-g1ac4c81/Win32/GL4.4 The map is present, but corrupted (same as in the original claim from @Sortale). See screenshot. |
|
|
Oh yeah, infolog. |
|
|
infolog2 shows no relevant errors, back to square one. until/unless someone donates ATI hardware I don't see this being solved. |
|
|
testing 104.0.1-296-gde4cee3 unchanged springsettings.cfg screenshot make garbage file [hilarious] (want me to upload it?) [still a png file just have garbage name] still have the tiling problem of 104.0.1-49-g70f081f.png infolog: https://pastebin.com/1H1pSJYS with "GLContextMajorVersion=4 GLContextMinorVersion=5" setting same problem no change pic https://imgur.com/a/eRWLT infolog https://pastebin.com/rR51AkKY what if I just have you control my computer? through team viewer or something? [massive privacy problem though] |
|
|
a couple of people with older gfx hardware experience this on 104.104... maintenance (not regular 104) so i hooked up my old Radeon HD5470 and it had it too! |
|
|
attached infolog.txt |
|
|
UsePBO = 0 did the trick to fix it |
|
|
Yes, I've been playing with UsePBO = 0 last couple of months. That said I previously used to play with default setting, which i believe UsePBO = 1, just fine. So at some point of time the regression happened, thus we have this ticket. Can anyone guide how performance hitting is UsePBO = 0 vs 1? |
|
|
I've actually figured this out. It wasn't a regression but the other way round. Previous windows glew didn't support PBOs at all so UsePBO had no effect (they were just not used). Once it did start supporting these, UsePBO started working and we've hit the issue at hand (which already happened in linux with newer glew versions). |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2017-10-21 09:56 | Sortale | New Issue | |
| 2017-10-21 10:03 | Sortale | File Added: 104.0.1-49-g70f081f.png | |
| 2017-10-21 10:07 | Sortale | File Added: 103.0.1-540-g9fb8b24.png | |
| 2017-10-21 13:40 | Kloot | Note Added: 0018542 | |
| 2017-10-22 20:05 | abma | Status | new => feedback |
| 2017-10-26 14:42 | Sortale | Note Added: 0018554 | |
| 2017-10-26 14:42 | Sortale | Status | feedback => new |
| 2017-10-26 14:54 | hokomoko | Note Added: 0018555 | |
| 2017-10-26 15:10 | hokomoko | Note Edited: 0018555 | |
| 2017-10-26 15:11 | hokomoko | Note Edited: 0018555 | |
| 2017-10-26 16:03 | Kloot | Note Added: 0018556 | |
| 2017-10-27 16:05 | Sortale | Note Added: 0018557 | |
| 2017-10-27 16:28 | hokomoko | Note Added: 0018558 | |
| 2017-10-28 05:02 | Google_Frog | Note Added: 0018559 | |
| 2017-10-28 12:44 | lhog | Note Added: 0018560 | |
| 2017-10-28 13:03 | lhog | Note Added: 0018561 | |
| 2017-10-28 13:06 | hokomoko | Note Added: 0018562 | |
| 2017-10-28 13:10 | lhog | Note Added: 0018563 | |
| 2017-10-28 13:39 | Kloot | Note Added: 0018564 | |
| 2017-10-28 13:50 | hokomoko | Note Added: 0018565 | |
| 2017-10-28 14:02 | Kloot | Note Edited: 0018564 | |
| 2017-10-28 16:45 | lhog | Note Added: 0018566 | |
| 2017-10-28 17:10 | hokomoko | Note Added: 0018567 | |
| 2017-10-28 17:15 | Kloot | Note Added: 0018568 | |
| 2017-10-28 19:34 | lhog | Note Added: 0018569 | |
| 2017-10-30 00:09 | Sortale | Note Added: 0018572 | |
| 2017-10-30 00:18 | Sortale | Note Added: 0018573 | |
| 2017-10-30 00:23 | hokomoko | Note Added: 0018574 | |
| 2017-10-30 00:25 | Sortale | Note Added: 0018575 | |
| 2017-10-30 03:28 | Google_Frog | Note Added: 0018576 | |
| 2017-10-30 04:05 | hokomoko | Note Added: 0018577 | |
| 2017-10-30 09:04 | Sortale | Note Added: 0018578 | |
| 2017-10-30 10:16 | hokomoko | Note Added: 0018579 | |
| 2017-10-30 11:32 | hokomoko | Note Added: 0018580 | |
| 2017-10-30 14:43 | lhog | File Added: screen00029.jpg | |
| 2017-10-30 14:43 | lhog | Note Added: 0018581 | |
| 2017-10-30 18:59 | Kloot | Note Added: 0018582 | |
| 2017-10-30 19:05 | Kloot | Note Edited: 0018582 | |
| 2017-10-30 19:23 | hokomoko | Note Added: 0018583 | |
| 2017-10-30 19:24 | hokomoko | Note Edited: 0018583 | |
| 2017-10-30 19:25 | hokomoko | Note Edited: 0018583 | |
| 2017-10-30 19:29 | Kloot | Note Added: 0018584 | |
| 2017-10-30 19:29 | Kloot | Note Edited: 0018584 | |
| 2017-10-30 19:31 | hokomoko | Note Added: 0018585 | |
| 2017-11-01 11:52 | Sortale | Note Added: 0018594 | |
| 2017-11-01 13:24 | Kloot | Note Edited: 0018584 | |
| 2017-11-01 13:24 | Kloot | Note Added: 0018595 | |
| 2017-11-01 13:25 | Kloot | Note Edited: 0018595 | |
| 2017-11-01 13:35 | Kloot | Note Edited: 0018595 | |
| 2017-11-01 14:50 | lhog | Note Added: 0018596 | |
| 2017-11-01 16:16 | Kloot | Note Added: 0018597 | |
| 2017-11-01 19:40 | lhog | File Added: infolog.txt | |
| 2017-11-01 19:40 | lhog | Note Added: 0018598 | |
| 2017-11-01 22:29 | Kloot | Note Added: 0018599 | |
| 2017-11-01 22:42 | Kloot | Note Edited: 0018599 | |
| 2017-11-02 13:17 | Kloot | Note Added: 0018603 | |
| 2017-11-02 14:16 | lhog | File Added: screen00030.jpg | |
| 2017-11-02 14:16 | lhog | Note Added: 0018604 | |
| 2017-11-02 14:17 | lhog | File Added: infolog-2.txt | |
| 2017-11-02 14:17 | lhog | Note Added: 0018605 | |
| 2017-11-02 19:02 | Kloot | Note Added: 0018606 | |
| 2017-11-03 07:33 | Sortale | Note Added: 0018607 | |
| 2017-12-08 02:28 | Floris | File Added: screen00016.jpg | |
| 2017-12-08 02:28 | Floris | Note Added: 0018676 | |
| 2017-12-08 02:35 | Floris | File Added: infolog-3.txt | |
| 2017-12-08 02:35 | Floris | Note Added: 0018677 | |
| 2017-12-08 02:45 | Floris | Note Added: 0018678 | |
| 2017-12-08 12:54 | lhog | Note Added: 0018679 | |
| 2017-12-08 13:00 | hokomoko | Note Added: 0018680 | |
| 2019-04-11 00:42 | Kloot | Assigned To | => Kloot |
| 2019-04-11 00:42 | Kloot | Status | new => resolved |
| 2019-04-11 00:42 | Kloot | Resolution | open => fixed |
| 2019-04-11 00:42 | Kloot | Fixed in Version | => 104.0 +git |
