#422 Progress of the 3.7.0 patch

Closed
opened 11 months ago by Krock · 44 comments
Krock commented 11 months ago

Official release date: 24 May (pre-download 22 Apr)

Patch completion/script unlock: 27 May ¹

Approximate unpatched/vanilla compatibility: 3 Jun ²

Progress so far (updated incrementally):

  • DONE - Locate the "check functions"
  • DONE - Locate the service start function
  • DONE - Locate & patch the manufacturer string ("Wine", new method)
  • DONE - Write patch skeleton
  • DONE - Anti-abuse quirk
  • DONE - Check for new launcher key
  • TODO - Secondary "low CPU" patch
    • DLL got statically linked into UserAssembly.dll. This is a surprising change; I wonder how they came to this decision. Unfortunately server-sent commands (including the "low CPU" cause) will now be more difficult/tedious to analyze.
    • OS and CN part ready

Tasks after official launch:

  • DONE - Dump "correct" values from a Windows installation
  • DONE - Check for newly added logging servers
  • DONE - Publish patch for testing
  • TODO - Test that patch (crowdsourcing)

Note: The CN counterpart is done by @y0soro and will have the same changes.

Please only leave a comment if you have questions or helpful inputs. Note that development takes place in the doctor branch until there is a patch available for testing.

¹) Might also take longer. New code is an unknown time factor.

²) Likely determined by servers. See issue #404.

Official release date: 24 May (pre-download 22 Apr) Patch completion/script unlock: 27 May ¹ Approximate unpatched/vanilla compatibility: 3 Jun ² **Progress so far (updated incrementally):** * DONE - Locate the "check functions" * DONE - Locate the service start function * DONE - Locate & patch the manufacturer string ("Wine", new method) * DONE - Write patch skeleton * DONE - Anti-abuse quirk * DONE - Check for new launcher key * TODO - Secondary "low CPU" patch * DLL got statically linked into UserAssembly.dll. This is a surprising change; I wonder how they came to this decision. Unfortunately server-sent commands (including the "low CPU" cause) will now be more difficult/tedious to analyze. * OS and CN part ready **Tasks after official launch:** * DONE - Dump "correct" values from a Windows installation * DONE - Check for newly added logging servers * DONE - Publish patch for testing * TODO - Test that patch (crowdsourcing) Note: The CN counterpart is done by @y0soro and will have the same changes. Please only leave a comment if you have questions or helpful inputs. Note that development takes place in the `doctor` branch until there is a patch available for testing. ¹) Might also take longer. New code is an unknown time factor. ²) Likely determined by servers. See issue #404.
y0soro commented 11 months ago

Hello CN players,

The CN patch would be postponed for a few days due to personal business.

Hello CN players, The CN patch would be postponed for a few days due to personal business.
Krock commented 11 months ago
Owner

Unfortunately it seems that I have lost control about the xlua patch required after approx. Friday to disable the CPU-hungry mhypbase.dll.

Disabling mhypbase.dll is still desired and I will try to provide a patch later. In the meantime, you might try to rename the file until the game crashes upon login.

On my slightly dated system, disabling mhypbase resulted in the following CPU usage change:

  • 50% -> 30% @ Main menu (also on pcgoogle)
  • 80% -> 60% @ Start area
Unfortunately it seems that I have lost control about the xlua patch required after approx. Friday to disable the CPU-hungry mhypbase.dll. Disabling mhypbase.dll is still desired and I will try to provide a patch later. In the meantime, you might try to rename the file until the game crashes upon login. On my slightly dated system, disabling mhypbase resulted in the following CPU usage change: * 50% -> 30% @ Main menu (also on pcgoogle) * 80% -> 60% @ Start area
Krock commented 11 months ago
Owner

Main patch for 3.6.0 (OS) is done

As usual, please use only testing accounts until the patch is confirmed working. If everything goes well, patch unlock will be on 27 May.

The unpatched/vanilla installation did currently not work on my side. If the pattern continues, it should work after approx. 3 Jun. If you have any information on this matter, please let me know so that the setup instructions can be updated accordingly.

Thank you.


EDIT: Depending on my free time, I might provide a CN patch in the next 48 hours.


The crash fix for renaming mhypbase will follow later if necessary.

#### Main patch for 3.6.0 (OS) is done As usual, please **use only testing accounts** until the patch is confirmed working. If everything goes well, patch unlock will be on 27 May. The unpatched/vanilla installation did currently not work on my side. If the pattern continues, it should work after approx. 3 Jun. If you have any information on this matter, please let me know so that the setup instructions can be updated accordingly. Thank you. --- EDIT: Depending on my free time, I might provide a CN patch in the next 48 hours. --- The crash fix for renaming mhypbase will follow later if necessary.

On startup a message popup appears (see attachment) and after it disappears nothing happens. Even ui buttons (on the sides of the screen) not working.

I tried to reapply patch, renaming mhypbase.dll, but it's not working.

Edit: Renaming file "Genshin Impact/GenshinImpact_Data/Plugins/telemetry.dll" to "Telemetry.dll" solved the issue

On startup a message popup appears (see attachment) and after it disappears nothing happens. Even ui buttons (on the sides of the screen) not working. I tried to reapply patch, renaming mhypbase.dll, but it's not working. Edit: Renaming file "Genshin Impact/GenshinImpact_Data/Plugins/telemetry.dll" to "Telemetry.dll" solved the issue
There is no content yet.
Ph42oN commented 11 months ago

For me on first try it stayed on the same screen others above reported, but without any message. Next try i got login screen, but i cannot log in, when i try it says "Connection failed. Try again later."

There is some things changed in this patch that don't usually change. FPS unlocker needed update while it normally doesn't with new patch.

Edit: We found fix for this issue on AAGL discord. Launcher downloaded new telemetry.dll, and old was still there as Telemetry.dll. Game tried to load old version and login does not work with that. To fix this issue delete Telemetry.dll.

For me on first try it stayed on the same screen others above reported, but without any message. Next try i got login screen, but i cannot log in, when i try it says "Connection failed. Try again later." There is some things changed in this patch that don't usually change. FPS unlocker needed update while it normally doesn't with new patch. Edit: We found fix for this issue on AAGL discord. Launcher downloaded new telemetry.dll, and old was still there as Telemetry.dll. Game tried to load old version and login does not work with that. To fix this issue delete Telemetry.dll.

ok, this file should be removed

GenshinImpact_Data/Plugins/Telemetry.dll

than it works

From AAGL discord:

the new function is placed inside the telemetry.dll and we have Telemetry.dll from the old version it seems the game load the old one

ok, this file should be removed ``` GenshinImpact_Data/Plugins/Telemetry.dll ``` than it works From AAGL discord: > the new function is placed inside the ```telemetry.dll``` and we have ```Telemetry.dll``` from the old version it seems the game load the old one

if you are using aagl launcher to update then navigate to Genshin Impact Game/GenshinImpact_Data/Plugins/ and delete the file Telemetry.dll and rename the file telemetry.dll to Telemetry.dll

it seems they changed the file name from uppercase to lowercase and on Linux system, the file will not be overwritten which cause the issue.

EDIT: just delete the old Telemetry.dll will do.

if you are using aagl launcher to update then navigate to Genshin Impact Game/GenshinImpact_Data/Plugins/ and delete the file ```Telemetry.dll``` and rename the file ```telemetry.dll``` to ```Telemetry.dll``` it seems they changed the file name from uppercase to lowercase and on Linux system, the file will not be overwritten which cause the issue. EDIT: just delete the old ```Telemetry.dll``` will do.

Currently playing, ~1hr 30m in. So far no issues encountered.

Thank you very much!


(I can't re-confirm this yet until the game change the kernel driver used)

Based on my previous testing, it appears the CPU usage problem is somewhat correlated to the kernel driver used. when it loaded mhyprot, the CPU usage went up, while when it loaded, hoyoprotect the CPU usage behaved normally. I'll try to test it again once the server changes kernel driver to see if it's still the case. If it's still the case, it could be an "alternative" to patching base dll (Well, if it doesn't swap randomly), though it's server-side bound.

Currently playing, ~1hr 30m in. So far no issues encountered. Thank you very much! -------------- (I can't re-confirm this yet until the game change the kernel driver used) Based on my previous testing, it appears the CPU usage problem is somewhat correlated to the kernel driver used. when it loaded mhyprot, the CPU usage went up, while when it loaded, hoyoprotect the CPU usage behaved normally. I'll try to test it again once the server changes kernel driver to see if it's still the case. If it's still the case, it could be an "alternative" to patching base dll (Well, if it doesn't swap randomly), though it's server-side bound.

@Krock can I ask you to add Telemetry.dll file removing to the patching script as this issue is related to the whole linux community and fix should be done on the patch level

@Krock can I ask you to add `Telemetry.dll` file removing to the patching script as this issue is related to the whole linux community and fix should be done on the patch level
liuk commented 11 months ago

i'm having a hard time trying to update it

i'm having a hard time trying to update it
brpt commented 11 months ago

Here's the hashes for the conflicting files in case it helps anyone:

$ md5sum GenshinImpact_Data/Plugins/?elemetry.dll 
70e2da50e63ca6a3c01d48cff6958b83  GenshinImpact_Data/Plugins/telemetry.dll
698fe27d7945137c5a36b4fe6dd35c83  GenshinImpact_Data/Plugins/Telemetry.dll

I deleted Telemetry.dll and was able to log in.

Here's the hashes for the conflicting files in case it helps anyone: ``` $ md5sum GenshinImpact_Data/Plugins/?elemetry.dll 70e2da50e63ca6a3c01d48cff6958b83 GenshinImpact_Data/Plugins/telemetry.dll 698fe27d7945137c5a36b4fe6dd35c83 GenshinImpact_Data/Plugins/Telemetry.dll ``` I deleted `Telemetry.dll` and was able to log in.

The game doesn't launch with wine-staging for me with both the official launcher and launcher.bat. Switching to wine package (Arch Linux), lets me launch the game. Loss of input issue is annoying but virtual desktop resolves that.

  • Arch Linux (6.3.3-zen)
  • NVIDIA 530.41.03
  • Wine 8.8
  • DXVK 2.2

Thanks!

edit: Can confirm renaming mhypbase.dll fixes the game not launching with wine-staging as with @Ennea,

The game doesn't launch with `wine-staging` for me with both the official launcher and `launcher.bat`. Switching to `wine` package (Arch Linux), lets me launch the game. Loss of input issue is annoying but virtual desktop resolves that. - Arch Linux (6.3.3-zen) - NVIDIA 530.41.03 - Wine 8.8 - DXVK 2.2 Thanks! edit: Can confirm renaming mhypbase.dll fixes the game not launching with wine-staging as with @Ennea,

Everything works, but HUGE CPU USAGE. Increased from ~15-20% to ~50%.

It will burn T_T

OS: Arch Linux x86_64
Kernel: 6.3.3-zen1-1-zen
CPU: AMD Ryzen 5 5600X
GPU: AMD ATI Radeon RX 6500 XT
Wine: GE-Proton8-4
DXVK: 2.2

https://www.youtube.com/watch?v=y8kVDKCvPyQ IMAGE ALT TEXT HERE

Everything works, but HUGE CPU USAGE. Increased from ~15-20% to ~50%. It will burn T_T OS: Arch Linux x86_64 Kernel: 6.3.3-zen1-1-zen CPU: AMD Ryzen 5 5600X GPU: AMD ATI Radeon RX 6500 XT Wine: GE-Proton8-4 DXVK: 2.2 https://www.youtube.com/watch?v=y8kVDKCvPyQ [![IMAGE ALT TEXT HERE](https://sun9-6.userapi.com/aLMQ2xfcCqx1ZKok3J6L8Oj-S1v3Hz0I3gy7Sw/SbPwnufteUk.jpg)](https://www.youtube.com/watch?v=y8kVDKCvPyQ)

@Cha14ka If your Wine/Proton is set to fsync, try switching to esync. That should return to normal values (Only works with Zen/1000 Hz kernels).

@Cha14ka If your Wine/Proton is set to fsync, try switching to esync. That should return to normal values (Only works with Zen/1000 Hz kernels).

@pureshores Wow, that really helped. Thank you so much.

@pureshores Wow, that really helped. Thank you so much.
Ennea commented 11 months ago

Facing the same issue as @rodwood_tide, game does not launch with wine-staging 8.8. There's not even any errors as far as I can tell, it just.. dies.

Edit: Ah, renaming mhypbase.dll helps that..

Facing the same issue as @rodwood_tide, game does not launch with wine-staging 8.8. There's not even any errors as far as I can tell, it just.. dies. Edit: Ah, renaming mhypbase.dll helps that..
Scla commented 11 months ago

Can confirm the issue with logging in and the duplicated telemetry dll. As @Ph42oN said, deleting Telemetry.dll fixed the issue. Nothing else to report. Thanks for the good work!

Can confirm the issue with logging in and the duplicated telemetry dll. As @Ph42oN said, deleting `Telemetry.dll` fixed the issue. Nothing else to report. Thanks for the good work!
Krock commented 11 months ago
Owner

As of 948ccb3, the duplicated Telemetry.dll file will now be removed. I suggest to file an issue for the 3rd party launchers to perform case-insensitive file replacements upon update - update_gi.sh is likely affected by this issue too.


Unfortunately requiring a custom kernel with 1000 Hz timers is not a good long-term solution. I will try to find alternatives but only the "test in production" will show whether they avoid the crash.

As of 948ccb3, the duplicated Telemetry.dll file will now be removed. I suggest to file an issue for the 3rd party launchers to perform case-insensitive file replacements upon update - update_gi.sh is likely affected by this issue too. --- Unfortunately requiring a custom kernel with 1000 Hz timers is not a good long-term solution. I will try to find alternatives but only the "test in production" will show whether they avoid the crash.

I want to ask, have any players from China tried this patch?

I want to ask, have any players from China tried this patch?

@justin_12138 Didn't try the patch on the China edition since the CN patch is not released yet, but Global edition is working fine with the patch.

The renamed mhypbase.dll method according to #404 doesn't work for me.

I'm using AAGL BTW.

@justin_12138 Didn't try the patch on the China edition since the CN patch is not released yet, but Global edition is working fine with the patch. `The renamed mhypbase.dll method` according to https://notabug.org/Krock/dawn/issues/404#issuecomment-36308 doesn't work for me. I'm using AAGL BTW.
infine commented 11 months ago

Game would not launch on wine staging 8.7 (arch). Just fails silently (at least with no additional winedebug keys). Fixed by renaming mhypbase.dll

Got the Telemetry.dll issue too, not using 3rd party launchers, so update_gi.sh is also affected.

Game would not launch on wine staging 8.7 (arch). Just fails silently (at least with no additional winedebug keys). Fixed by renaming mhypbase.dll Got the Telemetry.dll issue too, not using 3rd party launchers, so update_gi.sh is also affected.

@justin_12138

UPDATE:

I tried to apply this global patch to CN edition in AAGL, it fails, then I disable the main patch in AAGL's settings, rename mhypbase.dll, now the CN edition works fine.

@justin_12138 UPDATE: I tried to apply this global patch to CN edition in AAGL, it fails, then I disable the main patch in AAGL's settings, rename `mhypbase.dll`, now the CN edition works fine.

it works well for CN edition!!!

it works well for CN edition!!!
bionux commented 11 months ago

@justin_12138

Are you using the official genshin launcher? I newly downloaded genshin today to update the patch but failed

@justin_12138 Are you using the official genshin launcher? I newly downloaded genshin today to update the patch but failed

@bionux delete line 16 in 370/patch.sh since it is in test status, and apply the patch again, no need to use 370/patch_anti_logincrash.sh

@bionux delete line 16 in 370/patch.sh since it is in test status, and apply the patch again, no need to use 370/patch_anti_logincrash.sh
bionux commented 11 months ago

@codychan

Yeah it's Ok ! Thank you very much but i don't konw why I can't find where we are taught to do this Is this always the case when downloading genshin in the future?

@codychan Yeah it's Ok ! Thank you very much but i don't konw why I can't find where we are taught to do this Is this always the case when downloading genshin in the future?

@bionux When you apply the patch, you can see the error message, search this message in the patch directory using some tool like rg/grep, then edit the files according to the message.

This is just a message saying this patch is in test status, you have to manually edit the file to make the patch really work, after the patch passes the test status, the author will remove those lines so we don't have to edit the files by ourselves.

@bionux When you apply the patch, you can see the error message, search this message in the patch directory using some tool like rg/grep, then edit the files according to the message. This is just a message saying this patch is in test status, you have to manually edit the file to make the patch really work, after the patch passes the test status, the author will remove those lines so we don't have to edit the files by ourselves.
liuk commented 11 months ago

mine is working fine by now, played a couple hours with no problem, and I'm using wine-staging on arch

BUT the paimon launcher still doesn't detect 3.7 versions HDiff patches for some reason

mine is working fine by now, played a couple hours with no problem, and I'm using wine-staging on arch BUT the paimon launcher still doesn't detect 3.7 versions HDiff patches for some reason
ZAGON commented 11 months ago

Ive not kept up to date with the project overall, but what is this about unpatched compatibility?

(I know off topic but I didnt want to open an issue for something tiny)

Ive not kept up to date with the project overall, but what is this about unpatched compatibility? (I know off topic but I didnt want to open an issue for something tiny)
shenmo commented 11 months ago
shenmo@shenmo-PC:~$ ~/.local/share/net.winegame.client/runners/wine/winehq-staging-8.5-x64only-x86_64/bin/wine /media/shenmo/XDATA/Genshin\ Impact/Genshin\ Impact\ Game/YuanShen.exe 
002c:fixme:winediag:LdrInitializeThunk wine-staging 8.5 is a testing version containing experimental patches.
002c:fixme:winediag:LdrInitializeThunk Please mention your exact version when filing bug reports on winehq.org.
0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0120:fixme:heap:EnumSystemFirmwareTables (0x4649524d, 0000000000000000, 0)
0124:fixme:kernelbase:AppPolicyGetThreadInitializationType FFFFFFFFFFFFFFFA, 000000000450FE10
0120:fixme:toolhelp:CreateToolhelp32Snapshot Unimplemented: heap list snapshot
0120:fixme:toolhelp:Heap32ListFirst : stub
0120:fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_PERFORMANCE_INFORMATION

CN Server. Applied the patch. 3.7

```bash shenmo@shenmo-PC:~$ ~/.local/share/net.winegame.client/runners/wine/winehq-staging-8.5-x64only-x86_64/bin/wine /media/shenmo/XDATA/Genshin\ Impact/Genshin\ Impact\ Game/YuanShen.exe 002c:fixme:winediag:LdrInitializeThunk wine-staging 8.5 is a testing version containing experimental patches. 002c:fixme:winediag:LdrInitializeThunk Please mention your exact version when filing bug reports on winehq.org. 0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005 0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005 0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005 0080:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005 0120:fixme:heap:EnumSystemFirmwareTables (0x4649524d, 0000000000000000, 0) 0124:fixme:kernelbase:AppPolicyGetThreadInitializationType FFFFFFFFFFFFFFFA, 000000000450FE10 0120:fixme:toolhelp:CreateToolhelp32Snapshot Unimplemented: heap list snapshot 0120:fixme:toolhelp:Heap32ListFirst : stub 0120:fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_PERFORMANCE_INFORMATION ``` CN Server. Applied the patch. 3.7
Alex72 commented 11 months ago

Played for a few hours yesterday, no problems as usual. Wine-Staging 8.7 + DXVK 2.1, NVidia 530.41.03

@ZAGON Game may be playable without any patches, just remove/rename mhypbase.dll in the game folder. Your mileage may vary tho, people report login error or game crashes.

Played for a few hours yesterday, no problems as usual. Wine-Staging 8.7 + DXVK 2.1, NVidia 530.41.03 @ZAGON Game may be playable without any patches, just remove/rename `mhypbase.dll` in the game folder. Your mileage may vary tho, people report login error or game crashes.
Krock commented 11 months ago
Owner

My time was quite limited in the past few days. So far I did find a way to properly prevent mhypbase from loading (without renaming the file). Surprisingly the server sided checks are - as of now - yet not active to check for functional differences. Usually this is due on Friday (local server time).

Thus a proper "low CPU" patch will have to wait until either:

  1. renaming mhypbase.dll proves to be reliable over the next few days or ...
  2. (in case of crashes) I can find a new workaround to prevent the responsible bytecode from being executed.

If you feel lucky and would like to reduce your CPU usage, you might give the file renaming trick a try - at the cost of risking a game crash upon login.

I am sorry for the inconvenience. The normal patch will be handled as usual, with a status update in the next 24 hours.

My time was quite limited in the past few days. So far I did find a way to properly prevent mhypbase from loading (without renaming the file). Surprisingly the server sided checks are - as of now - yet not active to check for functional differences. Usually this is due on Friday (local server time). Thus a proper "low CPU" patch will have to wait until either: 1. renaming mhypbase.dll proves to be reliable over the next few days or ... 2. (in case of crashes) I can find a new workaround to prevent the responsible bytecode from being executed. If you feel lucky and would like to reduce your CPU usage, you might give the file renaming trick a try - at the cost of risking a game crash upon login. I am sorry for the inconvenience. The normal patch will be handled as usual, with a status update in the next 24 hours.
Krock commented 11 months ago
Owner

Thank you for all test reports. The patch script is now unlocked.

I am keeping this issue open for potential crash reports caused by renaming mhypbase until a solution for the high CPU usage is found (see comment above).

EDIT: This only affects crashes where the game terminates with the Wine log message err:virtual:virtual_setup_exception stack overflow.

Thank you for all test reports. The patch script is now unlocked. I am keeping this issue open for potential crash reports caused by renaming mhypbase until a solution for the high CPU usage is found (see comment above). EDIT: This only affects crashes where the game terminates with the Wine log message `err:virtual:virtual_setup_exception stack overflow`.
jingai commented 11 months ago

I see the language packs path changed in update-gi.sh. If we had used the script prior to the change to update the game, should we move files to a new path?

I see the language packs path changed in update-gi.sh. If we had used the script prior to the change to update the game, should we move files to a new path?
Krock commented 11 months ago
Owner

@jingai These variables were only used to detect installed language packs and update those accordingly. Game version 3.6.0 already moved those files, hence the rather diff archive size. You should not need to take any action because the files are already located there after 3.6.0, especially after 3.7.0.

If this contradicts your situation or if there are other issues with it, please open a new issue.

@jingai These variables were only used to detect installed language packs and update those accordingly. Game version 3.6.0 already moved those files, hence the rather diff archive size. You should not need to take any action because the files are already located there after 3.6.0, especially after 3.7.0. If this contradicts your situation or if there are other issues with it, please open a new issue.
ZAGON commented 11 months ago

thanks @Alex72 I can confirm it does run without patch, no high cpu usage either. strange. this is a far too easy way to null an anti-cheat

thanks @Alex72 I can confirm it does run without patch, no high cpu usage either. strange. this is a far too easy way to null an anti-cheat
cybik commented 11 months ago

For the record, I've been monitoring patchless.

The mhypbase.dll switcheroo still works right now. Will keep observing.

For the record, I've been monitoring patchless. The `mhypbase.dll` switcheroo still works right now. Will keep observing.

Can update_gi.sh be used without concern or is there any potential reason to avoid it? I see it was updated 5 days ago so i'm guessing it should be fine?

Can `update_gi.sh` be used without concern or is there any potential reason to avoid it? I see it was updated 5 days ago so i'm guessing it should be fine?

@Megalomaniak it should be fine as a way to update the game without needing to run the official launcher. I'd say it's probably no less safe than installing the patch itself.

@Megalomaniak it should be fine as a way to update the game without needing to run the official launcher. I'd say it's probably no less safe than installing the patch itself.

@scorlibrian aye, can confirm. Worked well.

@scorlibrian aye, can confirm. Worked well.
Ph42oN commented 11 months ago

For me, in 3.6 the high cpu usage issue happened on 1000hz kernel and it was causing bad performance. It went away with low cpu patch. But this time, so far its performing normally so it doesnt look like i would need that patch. Maybe what they changed also fixed what was causing high cpu usage for me.

Edit: Now i tried patchless on my laptop, cpu usage went down and it runs smoother. This is on 100hz kernel, will check this later on my desktop where im using 1000hz.

For me, in 3.6 the high cpu usage issue happened on 1000hz kernel and it was causing bad performance. It went away with low cpu patch. But this time, so far its performing normally so it doesnt look like i would need that patch. Maybe what they changed also fixed what was causing high cpu usage for me. Edit: Now i tried patchless on my laptop, cpu usage went down and it runs smoother. This is on 100hz kernel, will check this later on my desktop where im using 1000hz.
cybik commented 11 months ago

Patchless support is spinning back up, right on cue. sub_channel=0 looks janked for now, but =3 (Epic) and =6 (Google) seem to have better luck launching successfully.

Patchless support is spinning back up, right on cue. `sub_channel=0` looks janked for now, but `=3` (Epic) and `=6` (Google) seem to have better luck launching successfully.
Krock commented 10 months ago
Owner

Closing as solved (or postponed) by 0af1f96f32. Please check the commit message for details.

The patchless approach should be further discussed in #404, including suggestions.

Closing as solved (or postponed) by 0af1f96f32. Please check the commit message for details. The patchless approach should be further discussed in #404, including suggestions.
Sign in to join this conversation.
Loading...
Cancel
Save
There is no content yet.