This prompt is reported repeatedly and randomly from CN Server players using macOS.
Informations confirmed so far:
This is not a ban. Everything goes normal after one day/few days
It's likely that the server tagged player id and login device, given the following behaviors observed:
Account A on device A get this prompt, but account B can still log in on device A.
Account A can also log in on device B flawlessly.
Account A can also log in on mobile device.
All operations (include patching game, setting up wine) are performed in the same manner (since they are all managed by the same launcher program)
It's completely random, not reproducible. And so far only a few but not all players get this prompt.
It has nothing to do with game files. Re-install the game as well as wine changes nothing.
Telemetry APIs are definitely blocked. No exception.
This prompt is reported repeatedly and randomly from CN Server players using macOS.
Informations confirmed so far:
1. This is not a ban. Everything goes normal after one day/few days
2. It's likely that the server tagged player id and login device, given the following behaviors observed:
* Account A on device A get this prompt, but account B can still log in on device A.
* Account A can also log in on device B flawlessly.
* Account A can also log in on mobile device.
* All operations (include patching game, setting up wine) are performed in the same manner (since they are all managed by the same launcher program)
3. It's completely random, not reproducible. And so far only a few but not all players get this prompt.
4. It has nothing to do with game files. Re-install the game as well as wine changes nothing.
5. Telemetry APIs are __definitely__ blocked. No exception.
I met the same question today on Arch Linux.
The Behaviour has no difference:
- Suddenly can't log in on my Arch Linux
- Log in on my Windows laptop successfully
- Try to re-install the game but nothing happened
- I will try to log in next a few days
Side note: blocking the telemetry servers will only prevent the client from sending information to the central logging server. The game server does still receive certain information, see #293.
I do not know what the cause is, but there are a few options:
A rather new Wine bug that causes heap corruptions in places that are not immediately noticeable.
You are using 3rd party applications such as the FPS unlocker and ... :
A new kind of sanity check is being tested in production
Manual intervention from server/content moderators (which would explain the randomness)
@jackfiled Did this happen on the CN server or the OS (international) server?
@nvfku So far patch_anti_logincrash.sh solved more issues (noticeably lower CPU usage) than it caused (disabling another component of the game). It would be interesting to see whether this patch makes the difference.
Side note: blocking the telemetry servers will only prevent the client from sending information to the central logging server. The game server does still receive certain information, see #293.
I do not know what the cause is, but there are a few options:
* A rather new Wine bug that causes heap corruptions in places that are not immediately noticeable.
* You are using 3rd party applications such as the FPS unlocker and ... :
* A new kind of sanity check is being tested in production
* Manual intervention from server/content moderators (which would explain the randomness)
@jackfiled Did this happen on the CN server or the OS (international) server?
@nvfku So far `patch_anti_logincrash.sh` solved more issues (noticeably lower CPU usage) than it caused (disabling another component of the game). It would be interesting to see whether this patch makes the difference.
@Krock I can confirm no 3rd party applications are involved.
Update: now (almost) all CN players using M1/M2 get this prompt (I say almost, because there is no report of survival). But interestingly there seems to be no effect on Intel users (although the sample size is relatively small)
Update 2: seems disabling patch_anti_logincrash.sh do solve the problem. Confirmed that it‘s this method effective.
Update 3: Now Intel users fail to launch game without patch_anti_logincrash.sh. And some M1 users get game crashed after entering the door
@Krock I can confirm no 3rd party applications are involved.
__Update__: now (almost) all CN players using M1/M2 get this prompt (I say almost, because there is no report of survival). But interestingly there seems to be no effect on Intel users (although the sample size is relatively small)
__Update 2__: seems disabling `patch_anti_logincrash.sh` do solve the problem. Confirmed that it‘s this method effective.
__Update 3__: Now Intel users fail to launch game without `patch_anti_logincrash.sh`. And some M1 users get game crashed after entering the door
And I try to disable patch_anti_logincrash.sh, the game just doesn't launch. I need to disable mhypbase.dll by mv mhypbase.dll mhypbase.dll.bak then the game start, but the prompt shows again.
I try to delete the user data in register, but this prompt shows again. So I think the statement "account and device are tagged in the server" is true.
@Krock I use the CN server.
And I try to disable `patch_anti_logincrash.sh`, the game just doesn't launch. I need to disable `mhypbase.dll` by `mv mhypbase.dll mhypbase.dll.bak` then the game start, but the prompt shows again.
I try to delete the user data in register, but this prompt shows again. So I think the statement "account and device are tagged in the server" is true.
These are a few of the obvious differences - ignoring the entirely different CPU architecture in the first place - between the platforms that would allow filtering. However, I would like to not jump to any hasty conclusions because the entire M1/M2 support is still relatively new - about two weeks, to my knowledge.
@3Shain
If the M1/M2 vs Intel gap holds true (does this contradict with @jackfiled's report?) it could be explained either by empty WMI strings:
https://github.com/wine-mirror/wine/blob/wine-8.4/dlls/wbemprox/builtin.c#L3417-L3433
... and/or suspicious-looking platform information (search for `PROCESSOR_ARCHITECTURE_ARM64`)
https://github.com/wine-mirror/wine/blob/wine-8.4/programs/wineboot/wineboot.c#L684-L691
These are a few of the obvious differences - ignoring the entirely different CPU architecture in the first place - between the platforms that would allow filtering. However, I would like to not jump to any hasty conclusions because the entire M1/M2 support is still relatively new - about two weeks, to my knowledge.
AFAIK, the M1/M2 models are still running Intel version of Wine via Rosetta 2 (to be precise, Wine 7.7 with hacks by CodeWeavers).
I don't think ARM version of Wine works on macOS.
@Krock
AFAIK, the M1/M2 models are still running Intel version of Wine via Rosetta 2 (to be precise, Wine 7.7 with hacks by CodeWeavers).
I don't think ARM version of Wine works on macOS.
@wushangwei Thank you for your advice. But sadly, it didn't work for me.
Here is what I tried and noticed:
First I use the patch_revert.sh, patches are reverted but it tells me "--- Backup 'mhypbase.dll.bak' is older than the current game version. Ignoring."
Then I launch the game and succeed, but when I log in, the prompt shows again.
So I try to use the "repair game" in the game launcher, which I think is equal to using the patch_revert.sh script
Apply the patch.sh, launch the game but it just doesn't launch.
It seems that if I want to launch the game, I must disable mhypbase.dll, but this work is done in the patch_anti_logincrash.sh script. So I'd like to know whether the file mhypbase.dll is disabled in yours?
@wushangwei Thank you for your advice. But sadly, it didn't work for me.
Here is what I tried and noticed:
- First I use the `patch_revert.sh`, patches are reverted but it tells me "--- Backup 'mhypbase.dll.bak' is older than the current game version. Ignoring."
- Then I launch the game and succeed, but when I log in, the prompt shows again.
- So I try to use the "repair game" in the game launcher, which I think is equal to using the `patch_revert.sh` script
- Apply the `patch.sh`, launch the game but it just doesn't launch.
It seems that if I want to launch the game, I must disable `mhypbase.dll`, but this work is done in the `patch_anti_logincrash.sh` script. So I'd like to know whether the file `mhypbase.dll` is disabled in yours?
I got this error on SteamOS too. I am also a CN server user.
At first, I thought the game file was damaged. So, I re-copied the game on my other pc that works fine.
But after I did this, I got an new error, every time I tried to login the game, it just need to download an update about 10,000MB, and it will show a Code:-9907 磁碟空間不足，請清理後重試(in English: not enough disk space, please retry after cleanup) error immediately.
But I still have 115GB left of my drive.
I got this error on SteamOS too. I am also a CN server user.
At first, I thought the game file was damaged. So, I re-copied the game on my other pc that works fine.
But after I did this, I got an new error, every time I tried to login the game, it just need to download an update about 10,000MB, and it will show a `Code:-9907 磁碟空間不足，請清理後重試(in English: not enough disk space, please retry after cleanup)` error immediately.
But I still have 115GB left of my drive.
I am pretty sure the game is the latest version, I copied it from windows I just played wit.
No clue about why I was getting the update request.
The screenshot about -9907:
I am pretty sure the game is the latest version, I copied it from windows I just played wit.
No clue about why I was getting the update request.
The screenshot about -9907:
@chinanoahli This is a whole different issue. See TROUBLESHOOTING.md. After solving that problem with winecfg you should be able to update and play.
@Hy.S Even though if Rosetta 2 is used, WINEDEBUG=-all wine wmic cpu get Name and similar system information very likely return an empty string or information about the ARM processor itself.
@chinanoahli This is a whole different issue. See [TROUBLESHOOTING.md](src/master/TROUBLESHOOTING.md#an-error-code-appears). After solving that problem with `winecfg` you should be able to update and play.
---
@Hy.S Even though if Rosetta 2 is used, `WINEDEBUG=-all wine wmic cpu get Name` and similar system information very likely return an empty string or information about the ARM processor itself.
@Krockwine wmic cpu get Name gives VirtualApple on Rosetta. I managed to fake it (to a normal Intel) by modifying wine but it didn't help at all. (I'm sure the modification is effective, have checked wmic as well as regedit ,and all web api calls carry the faked cpu id. I don't know what will happen to game data packets (could be sniffed but all the tools are made for Windows). Anyway I suspect it has nothing to do with cpu info.
And I found this [Login] OnPlayerLoginRsp: 39 in a logging web request. Comparing to what a successful login gives: [WindSeed] AreaID: 102651, AreaType: 1
I want to host a PS to verify what information the client will sent through PlayerLoginReq. So far wasted several hours and make no progress :/
@Krock `wine wmic cpu get Name` gives `VirtualApple` on Rosetta. I managed to fake it (to a normal Intel) by modifying wine but it didn't help at all. (I'm sure the modification is effective, have checked `wmic` as well as `regedit` ,and all web api calls carry the faked cpu id. I don't know what will happen to game data packets (could be sniffed but all the tools are made for Windows). Anyway I suspect it has nothing to do with cpu info.
And I found this `[Login] OnPlayerLoginRsp: 39` in a logging web request. Comparing to what a successful login gives: `[WindSeed] AreaID: 102651, AreaType: 1`
---
I want to host a PS to verify what information the client will sent through `PlayerLoginReq`. So far wasted several hours and make no progress :/
Update 4: Now this prompt is 100% reproducible on Apple Silicon devices. And some of them successfully enter the game by disable patch_anti_logincrash.sh. Others either 1.the game doesn't run at all or 2.enter the door, see progress bar, then the game crashes.
Note: it (this issue) happens on CN server only. No report from OS server so far.
__Update 4__: Now this prompt is 100% reproducible on Apple Silicon devices. And some of them successfully enter the game by disable `patch_anti_logincrash.sh`. Others either 1.the game doesn't run at all or 2.enter the door, see progress bar, then the game crashes.
Note: it (this issue) happens on CN server only. No report from OS server so far.
@3Shain Without patch_anti_logincrash.sh game might crash if server sends WindSeed packet. It sends them arbitrarily and it may help to log in and play a few times on an Android / iOS device to lower your "security threat level".
@3Shain Without `patch_anti_logincrash.sh` game might crash if server sends WindSeed packet. It sends them arbitrarily and it may help to log in and play a few times on an Android / iOS device to lower your "security threat level".
@Krock Oh, looks like I miss this part, I read the doc in deck screen, It is too small.
-9907 solved. But now, I got 数据异常，请重新登录游戏 error back...
Update:
Just using patch.sh repatch the game without patch_anti_logincrash.sh, works for SteamOS (and all other Linux? I guess) now, I will try it again tomorrow on my deck.
Update 2:
This works for SteamOS, I can login the game with the client only patched by patch.sh.
@Krock Oh, looks like I miss this part, I read the doc in deck screen, It is too small.
-9907 solved. But now, I got `数据异常，请重新登录游戏` error back...
**Update**:
Just using `patch.sh` repatch the game without `patch_anti_logincrash.sh`, works for SteamOS (and all other Linux? I guess) now, I will try it again tomorrow on my deck.
**Update 2**:
This works for SteamOS, I can login the game with the client only patched by `patch.sh`.
It's related to the absence of mhypbase.dll. This file must not only exist at the filesystem location but also loaded and executed correctly.
I managed to make it loaded (so e.g. EnumProcessModules will indicate it's loaded on memory) but not executed at all (DllMain as well as Initialize are bypassed). The prompt still occurs.
It happens beforeWindSeedClientNotify. Not even be sent from the server.
But the problem is, mhypbase makes the game crash on several devices. (Even the same wine (binary) are used)
But why would the game could run flawlessly without mhypbase, I'm curiousand why patch_anti_logincrash need to disable mhypbase by renaming? cuz that's the whole point of the patch
I've gathered more information about this issue:
1. It's related to the absence of `mhypbase.dll`. This file __must__ not only exist at the filesystem location but also loaded and executed correctly.
* I managed to make it loaded (so e.g. `EnumProcessModules` will indicate it's loaded on memory) but not executed at all (`DllMain` as well as `Initialize` are bypassed). The prompt still occurs.
2. It happens __before__ `WindSeedClientNotify`. Not even be sent from the server.
3. But the problem is, `mhypbase` makes the game crash on several devices. (Even the same wine (binary) are used)
---
__But why would the game could run flawlessly without mhypbase, I'm curious__
<del>and why `patch_anti_logincrash` need to disable mhypbase by renaming?</del> cuz that's the whole point of the patch
Just to have it mentioned once: If device IDs are tagged, this attempt result might be slightly biased, causing false-negative testing results.
The server might as well return such error codes (39 = environment error) based on the data that is sent to it upon connect. Is there a errorCodeToPlatform value in that capture? That might help to narrow down the issue (except for 4302).
Is there anything helpful mentioned within the crash reports, e.g. WINEDEBUG output, output_log.txt and minidump files in %TEMP% ? Wine is far from perfect, so the crash might as well be caused by Wine itself.
EDIT: If you would like me to analyze log files of any sort to provide feedback, feel free to send them as an archive by email (see LICENSE.txt). Please keep in mind that network logs in particular might contain unique identifiers.
1. Just to have it mentioned once: If device IDs are tagged, this attempt result might be slightly biased, causing false-negative testing results.
2. The server might as well return such error codes (39 = environment error) based on the data that is sent to it upon connect. Is there a `errorCodeToPlatform` value in that capture? That might help to narrow down the issue (except for 4302).
3. Is there anything helpful mentioned within the crash reports, e.g. WINEDEBUG output, `output_log.txt` and minidump files in `%TEMP%` ? Wine is far from perfect, so the crash might as well be caused by Wine itself.
EDIT: If you would like me to analyze log files of any sort to provide feedback, feel free to send them as an archive by email (see LICENSE.txt). Please keep in mind that network logs in particular might contain unique identifiers.
Now I doubt it's the server tagging devices. Because you can see an immediate effect if you move away/put back mhypbase. And some player get this prompt on their first login.
39 is obtained from an error log sent via https. Unfortunately I failed to sniff the PlayerLogReq\Res packet for a failed login (which caused by this issue) although I did get such packet for a successful login.
The structure of PlayerLogReq:
{
"subChannelId": 1,
"languageType": 2,
"systemVersion": "Windows 10 (10.0.0) 64bit",
"deviceInfo": "operatingSystem:Windows 10 (10.0.0) 64bit&deviceModel:<device model (vendor), may leak info about mac>&graphicsDeviceName:AMD Radeon Pro 5300M&graphicsDeviceType:Direct3D11&graphicsDeviceVendor:ATI&graphicsDeviceVersion:Direct3D 11.0 [level 11.0]&graphicsMemorySize:4080&processorCount:12&processorFrequency:2600&processorType:Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz<shows 'VirtualApple @ 2.50GHz' on apple silicon devices>&systemMemorySize:16383",
"token": "<40digit hex>",
"deviceUuid": "<53digit hex>",
"deviceName": "<device name, may leak info about mac, while it's configurable>",
"clientVersion": "CNRELWin3.5.0_R13695448_S13586568_D13718257",
"platformType": 3,
"checksum": "50abc63416ccb9c814a462baf6242f722e459289a2db86e3bddeb8f98ee1cad227",
"checksumClientVersion": "CNRELWin3.5.0",
"environmentErrorCode": "<a base64 string ? no idea what's this>",
"accountType": 1,
"securityCmdReply": "<a base64 string>",
"clientDataVersion": 13718257,
"clientVerisonHash": "<a base64 string>",
"trackingIoInfo": {
"BHHPIADAKIM": "<The device model (The device vendor), may leak info about mac>"
}
}
(and nothing exciting in the response packet)
Another thing worth mentioning is that even though I can't sniff this packet for a failed login, I can still know the packet length, which is around 1500 bytes, but for a successful login it's about 800 bytes. Also the PlayerLoginRes has only 46 bytes (empty?), but normally it's much longer.
As for the crash, well maybe I should not say "crash". It's more like that the program terminate itself, without any helpful log (exitcode = 0, and I have tried various WINEDEBUG setting).
The method I used for packet sniffing. And that's why I confirm WindSeedClientNotify is not sent from server.
Now I doubt it's the server tagging devices. Because you can see an immediate effect if you move away/put back `mhypbase`. And some player get this prompt on their first login.
`39` is obtained from an error log sent via https. Unfortunately I failed to sniff the `PlayerLogReq\Res` packet for a failed login (which caused by this issue) although I did get such packet for a successful login.
The structure of `PlayerLogReq`:
```json
{
"subChannelId": 1,
"languageType": 2,
"systemVersion": "Windows 10 (10.0.0) 64bit",
"deviceInfo": "operatingSystem:Windows 10 (10.0.0) 64bit&deviceModel:<device model (vendor), may leak info about mac>&graphicsDeviceName:AMD Radeon Pro 5300M&graphicsDeviceType:Direct3D11&graphicsDeviceVendor:ATI&graphicsDeviceVersion:Direct3D 11.0 [level 11.0]&graphicsMemorySize:4080&processorCount:12&processorFrequency:2600&processorType:Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz<shows 'VirtualApple @ 2.50GHz' on apple silicon devices>&systemMemorySize:16383",
"token": "<40digit hex>",
"deviceUuid": "<53digit hex>",
"deviceName": "<device name, may leak info about mac, while it's configurable>",
"clientVersion": "CNRELWin3.5.0_R13695448_S13586568_D13718257",
"platformType": 3,
"checksum": "50abc63416ccb9c814a462baf6242f722e459289a2db86e3bddeb8f98ee1cad227",
"checksumClientVersion": "CNRELWin3.5.0",
"environmentErrorCode": "<a base64 string ? no idea what's this>",
"accountType": 1,
"securityCmdReply": "<a base64 string>",
"clientDataVersion": 13718257,
"clientVerisonHash": "<a base64 string>",
"trackingIoInfo": {
"BHHPIADAKIM": "<The device model (The device vendor), may leak info about mac>"
}
}
```
(and nothing exciting in the response packet)
Another thing worth mentioning is that even though I can't sniff this packet for a failed login, I can still know the packet length, which is around 1500 bytes, but for a successful login it's about 800 bytes. Also the `PlayerLoginRes` has only 46 bytes (empty?), but normally it's much longer.
As for the crash, well maybe I should not say "crash". It's more like that the program terminate itself, without any helpful log (exitcode = 0, and I have tried various WINEDEBUG setting).
---
The [method](https://sdl.moe/post/magic-sniffer/) I used for packet sniffing. And that's why I confirm `WindSeedClientNotify` is not sent from server.
Not only on arm, on my x86 archlinux pc also faced this issue. Reverting the patch and running patch.sh only solve it but will be stuck at update (if there is one). Running patch_anti_logincrash.sh will then show "Data error" again.
Not only on arm, on my x86 archlinux pc also faced this issue. Reverting the patch and running patch.sh only solve it but will be stuck at update (if there is one). Running patch_anti_logincrash.sh will then show "Data error" again.
As #404 goes, it seems that unpatched game can be run.
I try it and it works! Just start the official game launcher, run the "repair the game" to revert any patches and click "start game". And the prompt doesn't show!
As [#404](https://notabug.org/Krock/dawn/issues/404) goes, it seems that unpatched game can be run.
I try it and it works! Just start the official game launcher, run the "repair the game" to revert any patches and click "start game". And the prompt doesn't show!
I have met this problem today, on my archlinux with intel+nvidia. I cannot run the unpatched version. Disabling patch_anti_loginscreen.sh works for me.
I have met this problem today, on my archlinux with intel+nvidia. I cannot run the unpatched version. Disabling patch_anti_loginscreen.sh works for me.
Same here on Ubuntu 22.10 with wine-tkg 8.5, neither patched nor unpatched game can log in on the CN server.
It's weird that this doesn't happen a few days ago. Is my account gets noticed by the server somehow?
Same here on Ubuntu 22.10 with wine-tkg 8.5, neither patched nor unpatched game can log in on the CN server.
It's weird that this doesn't happen a few days ago. Is my account gets noticed by the server somehow?
@MATRIX-feather I have another Asia account, it runs well without issue on Arch Linux, only the CN account doesn't work (I can play the game with this CN account on my smartphone though).
@MATRIX-feather I have another Asia account, it runs well without issue on Arch Linux, only the CN account doesn't work (I can play the game with this CN account on my smartphone though).
Me too. This prompt comes again since the day before yesterday. I can still login and play on ps4, windows but not linux. I am CN server player too.
In this version, I only used the patch.sh without any other scripts. I also tried the latest script, after this prompt comes, I pull the script form here, and repatch the game, buy it did not work.
Me too. This prompt comes again since the day before yesterday. I can still login and play on ps4, windows but not linux. I am CN server player too.
In this version, I only used the `patch.sh` without any other scripts. I also tried the latest script, after this prompt comes, I pull the script form here, and repatch the game, buy it did not work.
As reported in #293, the server receives a list of the loaded system DLLs with an TRUST_E_NOSIGNATURE error code for each. This can be one way how they determine the "data error". However, I currently do not have the time to dissect the network and trace the information flow, nor could I reproduce this issue on my side for testing in the first place.
I am sorry for the inconvenience but currently I cannot provide a solution to this error.
(also in response to y0soro's comment)
As reported in #293, the server receives a list of the loaded system DLLs with an `TRUST_E_NOSIGNATURE` error code for each. This can be one way how they determine the "data error". However, I currently do not have the time to dissect the network and trace the information flow, nor could I reproduce this issue on my side for testing in the first place.
I am sorry for the inconvenience but currently I cannot provide a solution to this error.
On recently added observation of the error from this ticket, my own group and myself, I would say the current patching solutions are not sufficient to make CN game playable unless we have found a solution for this problem.
And due to my loss of interest in Genshin Impact, I am considering quitting the job of porting the patch to CN client.
I would probably still make a patch for 3.8 CN client if the error has not been observed after game release. But I can't guarantee for game versions after 3.8.
On recently added observation of the error from this ticket, my own group and myself, I would say the current patching solutions are not sufficient to make CN game playable unless we have found a solution for this problem.
And due to my loss of interest in Genshin Impact, I am considering quitting the job of porting the patch to CN client.
I would probably still make a patch for 3.8 CN client if the error has not been observed after game release. But I can't guarantee for game versions after 3.8.
Not that I want to side-track the discussion, but since 3.5 (3.4 if a rando redditor is to be believed), many OS players have been able to
play with one file rename (mhypbase.dll to something else) for a week and a half after an update lands
play completely unpatched starting around the Saturday of the week following the update (restore the mhypbase.dll file if/when the game stops starting)
In my case, since the 3.5 unpatched discovery, I've been
literally starting the upstream launcher through Steam and launching the game completely unpatched
starting the exe directly through steam for the week and a half following an update (3.6 and 3.7, same thing)
I'm wondering, has this unpatched behaviour been observed on the CN release?
(Also, we are literally at the door of 3.8 - this may apply again, or this way of playing may get invalidated. The next few hours will tell.)
Not that I want to side-track the discussion, but since 3.5 (3.4 if a rando redditor is to be believed), many OS players have been able to
* play with one file rename (`mhypbase.dll` to something else) for a week and a half after an update lands
* play completely unpatched starting around the Saturday of the week *following* the update (restore the `mhypbase.dll` file if/when the game stops starting)
In my case, since the 3.5 unpatched discovery, I've been
* literally starting the upstream launcher through Steam and launching the game completely unpatched
* starting the exe directly through steam for the week and a half following an update (3.6 and 3.7, same thing)
I'm wondering, has this unpatched behaviour been observed on the CN release?
(Also, we are literally at the door of 3.8 - this may apply again, or this way of playing may get invalidated. The next few hours will tell.)
@cybik I used to be able to launch the game with mhypbase.dll deleted after the 3.7 update for a long time. But one day (I don't remember the exact time) it asked me to login again after I switch wifi to network cable and then it showed the data error and I have been unable to play it since, even after I re-downloaded the game and patch it, it didn't work. I haven't tried the completely unpatched version yet.
@cybik I used to be able to launch the game with `mhypbase.dll` deleted after the 3.7 update for a long time. But one day (I don't remember the exact time) it asked me to login again after I switch wifi to network cable and then it showed the `data error` and I have been unable to play it since, even after I re-downloaded the game and patch it, it didn't work. I haven't tried the completely unpatched version yet.
I just logged in to 3.8 using Apple Game Porting Toolkit (wine 7.7) without any patch or modifications in CN server.
It seems some server-side checks are not enabled yet.
I just logged in to 3.8 using Apple Game Porting Toolkit (wine 7.7) without any patch or modifications in CN server.
It seems some server-side checks are not enabled yet.
Many of the players in the AAGL discord are reporting completely unpatched, untouched and unrenamed installs worked instantly. 3.8 might (I stress MIGHT) be the first unpatched-on-release version.
Many of the players in the AAGL discord are reporting completely unpatched, untouched and unrenamed installs worked instantly. 3.8 *might* (I stress MIGHT) be the first unpatched-on-release version.
@cybik Yeah, just tried, I can play this now, I hope this would be permanent, I haven't successfully logged in my CN account on my computer for over three weeks, otherwise I have to give up the game completely, I've always tired of login and doing the shitty tasks every day anyway.
@cybik Yeah, just tried, I can play this now, I hope this would be permanent, I haven't successfully logged in my CN account on my computer for over three weeks, otherwise I have to give up the game completely, I've always tired of login and doing the shitty tasks every day anyway.
Update: I completely removed all the game files, then install the official launcher and download the game file via the official launcher. The data error dialog doesn't popup anymore.
CN Server, data error again.
Update: I completely removed all the game files, then install the official launcher and download the game file via the official launcher. The data error dialog doesn't popup anymore.
Had tested on another archlinux installed notebook, and got the same error, so this maybe an Account Issue, I think.
I'm playing Genshin CN server, since ver3.8, I'd confirmed that I can play Genshin without any patch, just using wine to install, and use proton to start.
But yesterday I got a "Data Error" in the night.
But in morning, that didn't happen.
I told this to my friend, but he hadn't gotten any problem yesterday.
So I tried several ways to run the game, but nothing happened.
And just a moment ago, he told me he also got the same error message, and also cannot play.
We both tried to apply the 3.8 patch, and every way we can search from google or discord (such as An Anime Game Launcher team).
I think maybe we are just marked by the genshin server, because I see someone just change an account, and just can play well.
And we can still can play genshin on Windows or smartphone.
This maybe a CN server's special issue.
Had tested on another archlinux installed notebook, and got the same error, so this maybe an Account Issue, I think.
----------------
I'm playing Genshin CN server, since ver3.8, I'd confirmed that I can play Genshin without any patch, just using wine to install, and use proton to start.
But yesterday I got a "Data Error" in the night.
But in morning, that didn't happen.
I told this to my friend, but he hadn't gotten any problem yesterday.
So I tried several ways to run the game, but nothing happened.
And just a moment ago, he told me he also got the same error message, and also cannot play.
We both tried to apply the 3.8 patch, and every way we can search from google or discord (such as An Anime Game Launcher team).
I think maybe we are just marked by the genshin server, because I see someone just change an account, and just can play well.
And we can still can play genshin on Windows or smartphone.
This maybe a CN server's special issue.
I re-download the whole game, including the launcher. But it doesn't help. It's really confusing that the Bilibili server works while the CN server doesn't.
I re-download the whole game, including the launcher. But it doesn't help. It's really confusing that the Bilibili server works while the CN server doesn't.
CN server players, you don't need this patch anymore, you can just play the game with the original client with wine or proton directly. If you keek getting this error, you should delete your game client and then re-download it by the launcher.
Maybe the global game client will keep up soon. :)
There's a video in Bilibili BV1Ah411P7fK for a long time(2 months since this video was uploaded), I think the game company knows about this.
This issue ends now I think. Thanks to the developers who developed this patch, you guys are the best, even though I already left this game.
CN server players, you don't need this patch anymore, you can just play the game with the original client with wine or proton directly. If you keek getting this error, you should delete your game client and then re-download it by the launcher.
中国大陆服的玩家，现在原版游戏客户端已经可以直接在wine/proton里面跑了，不再需要补丁去阻止反作弊插件的运行了。4.0更新我已经在steam deck上测试过，可以直接跑。如果你没办法直接运行，请删除游戏客户端后，通过启动器重新下载游戏客户端本体（检查游戏完整性不一定能修复游戏）！
Maybe the global game client will keep up soon. :)
There's a video in [Bilibili BV1Ah411P7fK](https://b23.tv/cGUURun) for a long time(2 months since this video was uploaded), I think the game company knows about this.
This issue ends now I think. Thanks to the developers who developed this patch, you guys are the best, even though I already left this game.
We're just reporting the problem, if somebody else gets the same issue, then he/she can know that it is not his or her mistake.
我们不过是在报告这个问题，以便有其他人遇到同样问题时，可以确认这并非只是他/她的错误。
To chinanoahli :
Yes, we know that since 3.7 or earlier we can play Genshin CN version without any patch on linux.
当然，我们知道自从3.7或者更早版本，我们就已经可以在linux上无需补丁游玩原神中文服了。
But don't you know that it just works at the early-mid stage of every version, till the second gacha came, then we will get the "Data error" issue.
但你知道吗？这只在每个大版本的前中期有效，当每个版本的第二个池子出来时（只是根据经验判断大约是这个时期），就会出现“数据异常”的问题。
We're just reporting the problem, if somebody else gets the same issue, then he/she can know that it is not his or her mistake.
我们不过是在报告这个问题，以便有其他人遇到同样问题时，可以确认这并非只是他/她的错误。
Agree with kamikaze211, I got "Data error" again last week. Switch to win11 and it plays fine, same client file.
Some MacOS player also got same issue, it's not the player's own problem exactly
同意kamikaze211，我上周就已经出现“数据异常”了。切换到win11下可以正常玩，同样的客户端文件。
而且部分MacOS玩家也有遇到一样的情况，确实不是玩家自己的问题。
Agree with kamikaze211, I got "Data error" again last week. Switch to win11 and it plays fine, same client file.
Some MacOS player also got same issue, it's not the player's own problem exactly
@Alex72 The picture is from someone else, I personally use Steam Linux, only Proton Experimental/Proton 3~8/Proton Hotfix/Steam Linux Runtime available from Steam setting.
I tried Steam Linux Runtime, the game won't start.
I tried Proton Hotfix and Proton 8.0-3, it shows data error window, the same as the original one which is Proton Experimental
So what should I do if I use Steam?
@Alex72 The picture is from someone else, I personally use Steam Linux, only `Proton Experimental/Proton 3~8/Proton Hotfix/Steam Linux Runtime` available from Steam setting.
- I tried `Steam Linux Runtime`, the game won't start.
- I tried `Proton Hotfix` and `Proton 8.0-3`, it shows data error window, the same as the original one which is `Proton Experimental`
So what should I do if I use Steam?
I install wine-staging (v8.15-1) package from Arch Linux repo, and use its wine instead of the steam's proton to run the game, I get the same data error window.
I install a fresh game using wine-staging and run the game without any Steam components, I get the same data error window.
I guess my computer's device ID or this kind of info is put into GI's CN server's blacklist.
No other workaround or solution found so far.
@Alex72
- I install wine-staging (v8.15-1) package from Arch Linux repo, and use its wine instead of the steam's proton to run the game, I get the same data error window.
- I install a fresh game using wine-staging and run the game without any Steam components, I get the same data error window.
I guess my computer's device ID or this kind of info is put into GI's CN server's blacklist.
No other workaround or solution found so far.
Personal experience, this "data error" issue usually appears after I power off my steam deck while the game still running.
I got this issue 3 times from 3.7 and each time it appears when I forget to close the game first and power off/sleep my steam deck.
Just a guess, still not find out the root cause.
Personal experience, this "data error" issue usually appears after I power off my steam deck while the game still running.
I got this issue 3 times from 3.7 and each time it appears when I forget to close the game first and power off/sleep my steam deck.
Just a guess, still not find out the root cause.
It is more seems like that we are just be marked by the server.
Because that I had tried to change to another Linux installed pc,a completely installed game, but still get the "Data error" message.
However, I can play it on Windows or smart-phone or cloud-game.
@kavon
It is more seems like that we are just be marked by the server.
Because that I had tried to change to another Linux installed pc,a completely installed game, but still get the "Data error" message.
However, I can play it on Windows or smart-phone or cloud-game.
您的机器环境异常，请重启机器 error code: 10351-4001
got a new error from the 2nd banner of ver4.1
reproduce:
1. start game(cnrel)
2. normal gameplay for a few minutes
3. get the popup of "您的机器环境异常，请重启机器"
4. be exiled to the Gate
5. click the Gate to re-login
6. everything fine before closing the game.
The same as you, but the game is playable instead of "Data error" issue.
I also checked the same issue on AAGL at github, it seems to be a wine(proton) issue, just need to wait an update of wine/proton-ge/wine-ge.
@altioenb
The same as you, but the game is playable instead of "Data error" issue.
I also checked the same issue on AAGL at github, it seems to be a wine(proton) issue, just need to wait an update of wine/proton-ge/wine-ge.
I have played 4.2 version for 2 days, no longer to see this error.
But still cannot confirm if the "data error" will come back.
OK, after 3 days , I got this message "Error 10351-4001： 机器环境异常，请重启机器" again.
To CN server players:
It seems to be fixed by 4.2 version.
I have played 4.2 version for 2 days, no longer to see this error.
But still cannot confirm if the "data error" will come back.
OK, after 3 days , I got this message "Error 10351-4001： 机器环境异常，请重启机器" again.
I got infinite message "Error 10351-4001" when I played it a while.(This message would only pop up once before.) Seems the company tried to block Linux player again.
国服，登进去一会就弹窗（10351-4001），以前是只弹一次，现在次次弹。反正也是玩不了了。
I have tried:
DXVK 2.1/2.3
GE-Proton 8.25
Wine-GE-Proton 8.25/8.22/8.16
Lutris 7.2-2
(Yeah, I'm using AAGL)
CN player.
I got infinite message "Error 10351-4001" when I played it a while.(This message would only pop up once before.) Seems the company tried to block Linux player again.
国服，登进去一会就弹窗（10351-4001），以前是只弹一次，现在次次弹。反正也是玩不了了。
I have tried:
- DXVK 2.1/2.3
- GE-Proton 8.25
- Wine-GE-Proton 8.25/8.22/8.16
- Lutris 7.2-2
(Yeah, I'm using AAGL)
I got infinite message "Error 10351-4001" when I played it a while.(This message would only pop up once before.) Seems the company tried to block Linux player again.
Steam Deck, experimental compatibility layer, this always pops up after starting a game, and won't pop up again once you enter. This has been happening continuously until January 11, 2024.
The same with the Proton compatibility layer, I have been getting this non-stop since the 9th
The reason it popped up for me on the 11th is that I hadn't closed the game before. After a forced update and restarting the machine on the 11th, this kept popping up. The method mentioned in the thread about going into the game and then putting it into sleep mode for two minutes was personally tested and effective.
My environment is Debian testing with 8.0.1~repack-3 package with dxvk installed by winetricks. No other hacks, not patched game. I'm on CN server and I observed the identical issue :
Approximately before 2024/1/1 I received "Error 10351-4001" 1~2 minutes after entering game and had to re-enter. After re-enter I can play on.
On 2024/1/3 I got "Error 10351-4001" 1~2 minutes after every entrance which made the game unplayable.
Today I discovered this "login -> sleep -> resume" method, and tried:
Start game via wine explorer /desktop=genshin,3840x2160 YuanShen.exe in shell
After entering game, Ctrl-Z in shell to suspend the process
Wait about 7 minutes (maybe this can be shorter), use fg to resume the game
In the game, a network re-connection and reload is issued automatically.
After reloading, play on and no pop-ups anymore.
> CN player.
> I got infinite message "Error 10351-4001" when I played it a while.(This message would only pop up once before.) Seems the company tried to block Linux player again.
> 国服，登进去一会就弹窗（10351-4001），以前是只弹一次，现在次次弹。反正也是玩不了了。
Same observation as well.
I find a potential solution, that works for me at least. The originated post is in the comments of https://www.bilibili.com/video/BV1Lv411F7n1 :
> > > steamdeck，实验型兼容层，一直是启动游戏后弹出这个，再进就不会弹。直到2024年1月11日一直弹
> > 同proton兼容层，我是从9号开始不停弹的
> 我11号弹是因为之前没有关闭过游戏，11号强制更新重启机器后就不停跳这个了。楼里提到的进游戏后休眠两分钟的办法亲测有效
Translation
> > > Steam Deck, experimental compatibility layer, this always pops up after starting a game, and won't pop up again once you enter. This has been happening continuously until January 11, 2024.
> > The same with the Proton compatibility layer, I have been getting this non-stop since the 9th
> The reason it popped up for me on the 11th is that I hadn't closed the game before. After a forced update and restarting the machine on the 11th, this kept popping up. The method mentioned in the thread about going into the game and then putting it into sleep mode for two minutes was personally tested and effective.
My environment is Debian testing with 8.0.1~repack-3 package with dxvk installed by winetricks. No other hacks, not patched game. I'm on CN server and I observed the identical issue :
Approximately before 2024/1/1 I received "Error 10351-4001" 1~2 minutes after entering game and had to re-enter. After re-enter I can play on.
On 2024/1/3 I got "Error 10351-4001" 1~2 minutes after *every entrance* which made the game unplayable.
Today I discovered this "login -> sleep -> resume" method, and tried:
1. Start game via `wine explorer /desktop=genshin,3840x2160 YuanShen.exe` in shell
2. After entering game, Ctrl-Z in shell to suspend the process
3. Wait about 7 minutes (maybe this can be shorter), use `fg` to resume the game
4. In the game, a network re-connection and reload is issued automatically.
5. After reloading, play on and no pop-ups anymore.
A minor difference: after resume the game process, there's only network re-connection, while game re-loading did not occurred. Then I can play on with FPS unlocked
> 3. Wait about 7 minutes (maybe this can be shorter), use fg to resume the game
I tried another time. This time,
- I ran fps_unlocker.exe from https://github.com/34736384/genshin-fps-unlock
- I reduced the suspend time to 150s
A minor difference: after resume the game process, there's only network re-connection, while game re-loading did not occurred. Then I can play on with FPS unlocked
It seems that network re-connection is the way to escape the "Error 10351-4001". After entering the game, either disconnect the network or hang the game process can trigger a re-connection (no need to wait), and then I can play on.
> 另外我猜测更核心的是断网重连这一机制。原神 bug 千千万，断网重连占一半。
It seems that network re-connection is the way to escape the "Error 10351-4001". After entering the game, either disconnect the network or hang the game process can trigger a re-connection (no need to wait), and then I can play on.
10351-4001 happens again on . I experienced it from 2024-10-13. Re-entering, or kill -STOP; kill -CONT reconnecting does not help. Using /etc/hosts with
10351-4001 happens again on . I experienced it from 2024-10-13. Re-entering, or `kill -STOP; kill -CONT` reconnecting does not help. Using `/etc/hosts` with
```
0.0.0.0 public-data-api.mihoyo.com
0.0.0.0 devlog-upload.mihoyo.com
0.0.0.0 devlog-upload-os.hoyoverse.com
0.0.0.0 log-upload.mihoyo.com
0.0.0.0 log-upload-os.mihoyo.com
0.0.0.0 ys-log-upload-os.hoyoverse.com
0.0.0.0 sdk-log-upload-os.hoyoverse.com
0.0.0.0 ys-log-upload.mihoyo.com
0.0.0.0 uspider.yuanshen.com
0.0.0.0 dispatchcnglobal.yuanshen.com
```
does not help either.
According to discussions at https://github.com/yaagl/yet-another-anime-game-launcher/issues/466#issuecomment-2420308593, `-platform_type CLOUD_THIRD_PARTY_PC` can resolve the problem.
@littlewu2508 Thank you for the cross-reference. I did document the workaround as of 506b77f. Please let me know when the situation changes, or another workaround is needed.
@littlewu2508 Thank you for the cross-reference. I did document the workaround as of 506b77f. Please let me know when the situation changes, or another workaround is needed.
This prompt is reported repeatedly and randomly from CN Server players using macOS.
Informations confirmed so far:
I met the same question today on Arch Linux.
The Behaviour has no difference:
Suddenly can't log in on my Arch Linux
Log in on my Windows laptop successfully
Try to re-install the game but nothing happened
I will try to log in next a few days
i found the problem， i re-patched it，just only use patch.sh file，it ran! So i think the problem is about the file-patch_antilogincrash.sh
Side note: blocking the telemetry servers will only prevent the client from sending information to the central logging server. The game server does still receive certain information, see #293.
I do not know what the cause is, but there are a few options:
@jackfiled Did this happen on the CN server or the OS (international) server?
@nvfku So far
patch_anti_logincrash.shsolved more issues (noticeably lower CPU usage) than it caused (disabling another component of the game). It would be interesting to see whether this patch makes the difference.@Krock I can confirm no 3rd party applications are involved.
Update: now (almost) all CN players using M1/M2 get this prompt (I say almost, because there is no report of survival). But interestingly there seems to be no effect on Intel users (although the sample size is relatively small)
Update 2: seems disabling
patch_anti_logincrash.shdo solve the problem. Confirmed that it‘s this method effective.Update 3: Now Intel users fail to launch game without
patch_anti_logincrash.sh. And some M1 users get game crashed after entering the door@Krock I use the CN server.
And I try to disable
patch_anti_logincrash.sh, the game just doesn't launch. I need to disablemhypbase.dllbymv mhypbase.dll mhypbase.dll.bakthen the game start, but the prompt shows again.I try to delete the user data in register, but this prompt shows again. So I think the statement "account and device are tagged in the server" is true.
@3Shain
If the M1/M2 vs Intel gap holds true (does this contradict with @jackfiled's report?) it could be explained either by empty WMI strings:
https://github.com/wine-mirror/wine/blob/wine-8.4/dlls/wbemprox/builtin.c#L3417-L3433
... and/or suspicious-looking platform information (search for
PROCESSOR_ARCHITECTURE_ARM64)https://github.com/wine-mirror/wine/blob/wine-8.4/programs/wineboot/wineboot.c#L684-L691
These are a few of the obvious differences - ignoring the entirely different CPU architecture in the first place - between the platforms that would allow filtering. However, I would like to not jump to any hasty conclusions because the entire M1/M2 support is still relatively new - about two weeks, to my knowledge.
@Krock
AFAIK, the M1/M2 models are still running Intel version of Wine via Rosetta 2 (to be precise, Wine 7.7 with hacks by CodeWeavers).
I don't think ARM version of Wine works on macOS.
@jackfiled I'm also CN server user. I ran
patch_revert.shfirst to undo all the patches, then just ranpatch.sh, and I'm able to enter the game.@wushangwei Thank you for your advice. But sadly, it didn't work for me.
Here is what I tried and noticed:
patch_revert.sh, patches are reverted but it tells me "--- Backup 'mhypbase.dll.bak' is older than the current game version. Ignoring."patch_revert.shscriptpatch.sh, launch the game but it just doesn't launch.It seems that if I want to launch the game, I must disable
mhypbase.dll, but this work is done in thepatch_anti_logincrash.shscript. So I'd like to know whether the filemhypbase.dllis disabled in yours?I got this error on SteamOS too. I am also a CN server user.
At first, I thought the game file was damaged. So, I re-copied the game on my other pc that works fine.
But after I did this, I got an new error, every time I tried to login the game, it just need to download an update about 10,000MB, and it will show a
Code:-9907 磁碟空間不足，請清理後重試(in English: not enough disk space, please retry after cleanup)error immediately.But I still have 115GB left of my drive.
Same problem on WOA(Windows on Arm) device with
patch.shandpatch_anti_logincrash.shI am pretty sure the game is the latest version, I copied it from windows I just played wit.
No clue about why I was getting the update request.
The screenshot about -9907:
@chinanoahli This is a whole different issue. See TROUBLESHOOTING.md. After solving that problem with
winecfgyou should be able to update and play.@Hy.S Even though if Rosetta 2 is used,
WINEDEBUG=-all wine wmic cpu get Nameand similar system information very likely return an empty string or information about the ARM processor itself.@Krock
wine wmic cpu get NamegivesVirtualAppleon Rosetta. I managed to fake it (to a normal Intel) by modifying wine but it didn't help at all. (I'm sure the modification is effective, have checkedwmicas well asregedit,and all web api calls carry the faked cpu id. I don't know what will happen to game data packets (could be sniffed but all the tools are made for Windows). Anyway I suspect it has nothing to do with cpu info.And I found this
[Login] OnPlayerLoginRsp: 39in a logging web request. Comparing to what a successful login gives:[WindSeed] AreaID: 102651, AreaType: 1I want to host a PS to verify what information the client will sent through
PlayerLoginReq. So far wasted several hours and make no progress :/Update 4: Now this prompt is 100% reproducible on Apple Silicon devices. And some of them successfully enter the game by disable
patch_anti_logincrash.sh. Others either 1.the game doesn't run at all or 2.enter the door, see progress bar, then the game crashes.Note: it (this issue) happens on CN server only. No report from OS server so far.
@3Shain Without
patch_anti_logincrash.shgame might crash if server sends WindSeed packet. It sends them arbitrarily and it may help to log in and play a few times on an Android / iOS device to lower your "security threat level".@Krock Oh, looks like I miss this part, I read the doc in deck screen, It is too small.
-9907 solved. But now, I got
数据异常，请重新登录游戏error back...Update:
Just using
patch.shrepatch the game withoutpatch_anti_logincrash.sh, works for SteamOS (and all other Linux? I guess) now, I will try it again tomorrow on my deck.Update 2:
This works for SteamOS, I can login the game with the client only patched by
patch.sh.I've gathered more information about this issue:
It's related to the absence of
mhypbase.dll. This file must not only exist at the filesystem location but also loaded and executed correctly.EnumProcessModuleswill indicate it's loaded on memory) but not executed at all (DllMainas well asInitializeare bypassed). The prompt still occurs.It happens before
WindSeedClientNotify. Not even be sent from the server.But the problem is,
mhypbasemakes the game crash on several devices. (Even the same wine (binary) are used)But why would the game could run flawlessly without mhypbase, I'm curious
and whycuz that's the whole point of the patchpatch_anti_logincrashneed to disable mhypbase by renaming?errorCodeToPlatformvalue in that capture? That might help to narrow down the issue (except for 4302).output_log.txtand minidump files in%TEMP%? Wine is far from perfect, so the crash might as well be caused by Wine itself.EDIT: If you would like me to analyze log files of any sort to provide feedback, feel free to send them as an archive by email (see LICENSE.txt). Please keep in mind that network logs in particular might contain unique identifiers.
Now I doubt it's the server tagging devices. Because you can see an immediate effect if you move away/put back
mhypbase. And some player get this prompt on their first login.39is obtained from an error log sent via https. Unfortunately I failed to sniff thePlayerLogReq\Respacket for a failed login (which caused by this issue) although I did get such packet for a successful login.The structure of
PlayerLogReq:(and nothing exciting in the response packet)
Another thing worth mentioning is that even though I can't sniff this packet for a failed login, I can still know the packet length, which is around 1500 bytes, but for a successful login it's about 800 bytes. Also the
PlayerLoginReshas only 46 bytes (empty?), but normally it's much longer.As for the crash, well maybe I should not say "crash". It's more like that the program terminate itself, without any helpful log (exitcode = 0, and I have tried various WINEDEBUG setting).
The method I used for packet sniffing. And that's why I confirm
WindSeedClientNotifyis not sent from server.Not only on arm, on my x86 archlinux pc also faced this issue. Reverting the patch and running patch.sh only solve it but will be stuck at update (if there is one). Running patch_anti_logincrash.sh will then show "Data error" again.
As #404 goes, it seems that unpatched game can be run.
I try it and it works! Just start the official game launcher, run the "repair the game" to revert any patches and click "start game". And the prompt doesn't show!
I have met this problem today, on my archlinux with intel+nvidia. I cannot run the unpatched version. Disabling patch_anti_loginscreen.sh works for me.
Currently, neither no patch/patch.sh/both patch work on CN on my PC. Showing "31-4302" and "Data error, please log in again"
Same here on Ubuntu 22.10 with wine-tkg 8.5, neither patched nor unpatched game can log in on the CN server.
It's weird that this doesn't happen a few days ago. Is my account gets noticed by the server somehow?
@MATRIX-feather Is your account OK now? I got this issue yesterday, tried everything I know, but still couldn't make it work.
@codychan Yes, I can still play on my phone without restrictions
@MATRIX-feather I meant, can you play the game on your Ubuntu now?
@codychan Well, on Ubuntu I can't. It still displays "Data error, please log in again."
@MATRIX-feather I have another Asia account, it runs well without issue on Arch Linux, only the CN account doesn't work (I can play the game with this CN account on my smartphone though).
Me too. This prompt comes again since the day before yesterday. I can still login and play on ps4, windows but not linux. I am CN server player too.
In this version, I only used the
patch.shwithout any other scripts. I also tried the latest script, after this prompt comes, I pull the script form here, and repatch the game, buy it did not work.(also in response to y0soro's comment)
As reported in #293, the server receives a list of the loaded system DLLs with an
TRUST_E_NOSIGNATUREerror code for each. This can be one way how they determine the "data error". However, I currently do not have the time to dissect the network and trace the information flow, nor could I reproduce this issue on my side for testing in the first place.I am sorry for the inconvenience but currently I cannot provide a solution to this error.
On recently added observation of the error from this ticket, my own group and myself, I would say the current patching solutions are not sufficient to make CN game playable unless we have found a solution for this problem.
And due to my loss of interest in Genshin Impact, I am considering quitting the job of porting the patch to CN client.
I would probably still make a patch for 3.8 CN client if the error has not been observed after game release. But I can't guarantee for game versions after 3.8.
Not that I want to side-track the discussion, but since 3.5 (3.4 if a rando redditor is to be believed), many OS players have been able to
mhypbase.dllto something else) for a week and a half after an update landsmhypbase.dllfile if/when the game stops starting)In my case, since the 3.5 unpatched discovery, I've been
I'm wondering, has this unpatched behaviour been observed on the CN release?
(Also, we are literally at the door of 3.8 - this may apply again, or this way of playing may get invalidated. The next few hours will tell.)
@cybik I used to be able to launch the game with
mhypbase.dlldeleted after the 3.7 update for a long time. But one day (I don't remember the exact time) it asked me to login again after I switch wifi to network cable and then it showed thedata errorand I have been unable to play it since, even after I re-downloaded the game and patch it, it didn't work. I haven't tried the completely unpatched version yet.I just logged in to 3.8 using Apple Game Porting Toolkit (wine 7.7) without any patch or modifications in CN server. It seems some server-side checks are not enabled yet.
Many of the players in the AAGL discord are reporting completely unpatched, untouched and unrenamed installs worked instantly. 3.8 might (I stress MIGHT) be the first unpatched-on-release version.
@cybik Yeah, just tried, I can play this now, I hope this would be permanent, I haven't successfully logged in my CN account on my computer for over three weeks, otherwise I have to give up the game completely, I've always tired of login and doing the shitty tasks every day anyway.
it happened again, atleast i can play in the first week of each update
CN Server, data error again.
Update: I completely removed all the game files, then install the official launcher and download the game file via the official launcher. The data error dialog doesn't popup anymore.
Had tested on another archlinux installed notebook, and got the same error, so this maybe an Account Issue, I think.
I'm playing Genshin CN server, since ver3.8, I'd confirmed that I can play Genshin without any patch, just using wine to install, and use proton to start.
But yesterday I got a "Data Error" in the night.
But in morning, that didn't happen.
I told this to my friend, but he hadn't gotten any problem yesterday.
So I tried several ways to run the game, but nothing happened.
And just a moment ago, he told me he also got the same error message, and also cannot play.
We both tried to apply the 3.8 patch, and every way we can search from google or discord (such as An Anime Game Launcher team).
I think maybe we are just marked by the genshin server, because I see someone just change an account, and just can play well.
And we can still can play genshin on Windows or smartphone.
This maybe a CN server's special issue.
CNRel
I got 数据异常，请重新登录游戏 again now.
Data error, please log in again.window again, CN version, I'm playing it under Steam client without applying any patch.CN server, got
Data error, please log in again.starting from yesterday...I re-download the whole game, including the launcher. But it doesn't help. It's really confusing that the Bilibili server works while the CN server doesn't.
I can play v4.0 in Steam without any patch. Don't know when they will block it again.
I can also play today. It seems that this problem will disappear for a few days after every update.
I got a "Data error" today. It happened again.
CN server players, you don't need this patch anymore, you can just play the game with the original client with wine or proton directly. If you keek getting this error, you should delete your game client and then re-download it by the launcher.
中国大陆服的玩家，现在原版游戏客户端已经可以直接在wine/proton里面跑了，不再需要补丁去阻止反作弊插件的运行了。4.0更新我已经在steam deck上测试过，可以直接跑。如果你没办法直接运行，请删除游戏客户端后，通过启动器重新下载游戏客户端本体（检查游戏完整性不一定能修复游戏）！
Maybe the global game client will keep up soon. :)
There's a video in Bilibili BV1Ah411P7fK for a long time(2 months since this video was uploaded), I think the game company knows about this.
This issue ends now I think. Thanks to the developers who developed this patch, you guys are the best, even though I already left this game.
To chinanoahli :
Yes, we know that since 3.7 or earlier we can play Genshin CN version without any patch on linux.
当然，我们知道自从3.7或者更早版本，我们就已经可以在linux上无需补丁游玩原神中文服了。
But don't you know that it just works at the early-mid stage of every version, till the second gacha came, then we will get the "Data error" issue.
但你知道吗？这只在每个大版本的前中期有效，当每个版本的第二个池子出来时（只是根据经验判断大约是这个时期），就会出现“数据异常”的问题。
We're just reporting the problem, if somebody else gets the same issue, then he/she can know that it is not his or her mistake.
我们不过是在报告这个问题，以便有其他人遇到同样问题时，可以确认这并非只是他/她的错误。
同意kamikaze211，我上周就已经出现“数据异常”了。切换到win11下可以正常玩，同样的客户端文件。 而且部分MacOS玩家也有遇到一样的情况，确实不是玩家自己的问题。
Agree with kamikaze211, I got "Data error" again last week. Switch to win11 and it plays fine, same client file. Some MacOS player also got same issue, it's not the player's own problem exactly
"Data error, please log in again." window again, CN server here, time to give it up.
Someone else got this error window, and got an email saying the game is running on
Laptop (Wine(The Wine Project))@codychan You should switch to Wine-Staging instead of a vanilla Wine, it reports your motherboard data correctly.
@Alex72 The picture is from someone else, I personally use Steam Linux, only
Proton Experimental/Proton 3~8/Proton Hotfix/Steam Linux Runtimeavailable from Steam setting.Steam Linux Runtime, the game won't start.Proton HotfixandProton 8.0-3, it shows data error window, the same as the original one which isProton ExperimentalSo what should I do if I use Steam?
@codychan Honestly IDK, I do not use Steam, sorry...
@Alex72
I guess my computer's device ID or this kind of info is put into GI's CN server's blacklist.
No other workaround or solution found so far.
Personal experience, this "data error" issue usually appears after I power off my steam deck while the game still running. I got this issue 3 times from 3.7 and each time it appears when I forget to close the game first and power off/sleep my steam deck. Just a guess, still not find out the root cause.
@kavon
It is more seems like that we are just be marked by the server.
Because that I had tried to change to another Linux installed pc,a completely installed game, but still get the "Data error" message.
However, I can play it on Windows or smart-phone or cloud-game.
截止到2023年10月12日，游戏公告的「内容更新汇总」更新时间2023/10/12 17:20，你仍然可以不使用任何补丁，直接配合GE-proton8-4，直接使用原版游戏客户端进入游戏
只写中文了，懒得翻译成英语了，我只能说中国大陆服玩家都可以亲自试试，再说行不行
没试过就说不行的话。我也没什么好办法（
steam deck 系统版本 3.4.11(20231005.1) 核心版本 5.13.0-valve37-1-neptune
这个问题一般只会在每个版本的第20天才会开始出现，4.1版本截至目前为止不应该出现这个问题。如果下周一还是能正常进游戏，那么这个议题就可以关了。（反之也可以关了，因为就可以证明是米哈游服务端在人为管控）
@chinanoahli 原来还是你，可能你看不懂中文，英文也懒得看吧。 我们已经描述得很清楚了。 你可以在每个大版本第二个卡池出来后再试试。 如果你依然无法复现，说明你可能不是每天都玩。 总之，已经连续几个版本都有大量玩家在第二个卡池前后复现“数据异常”的问题了，而每个大版本的初期都是可以正常游戏的，这与设备无关。 目前能找到与异常相关的信息，只有服务器（大陆服独有）、账号（也许是每日登录账号）、时间（每个版本中后期）。
您的机器环境异常，请重启机器 error code: 10351-4001
got a new error from the 2nd banner of ver4.1
reproduce:
@altioenb
The same as you, but the game is playable instead of "Data error" issue. I also checked the same issue on AAGL at github, it seems to be a wine(proton) issue, just need to wait an update of wine/proton-ge/wine-ge.
To CN server players:
It seems to be fixed by 4.2 version.
I have played 4.2 version for 2 days, no longer to see this error.
But still cannot confirm if the "data error" will come back.
OK, after 3 days , I got this message "Error 10351-4001： 机器环境异常，请重启机器" again.
CN player.
I got infinite message "Error 10351-4001" when I played it a while.(This message would only pop up once before.) Seems the company tried to block Linux player again.
国服，登进去一会就弹窗（10351-4001），以前是只弹一次，现在次次弹。反正也是玩不了了。
I have tried:
(Yeah, I'm using AAGL)
Just use a dual boot for a while, and wait for developers to have more analysis and research.
Same observation as well.
I find a potential solution, that works for me at least. The originated post is in the comments of https://www.bilibili.com/video/BV1Lv411F7n1 :
Translation
My environment is Debian testing with 8.0.1~repack-3 package with dxvk installed by winetricks. No other hacks, not patched game. I'm on CN server and I observed the identical issue :
Approximately before 2024/1/1 I received "Error 10351-4001" 1~2 minutes after entering game and had to re-enter. After re-enter I can play on.
On 2024/1/3 I got "Error 10351-4001" 1~2 minutes after every entrance which made the game unplayable.
Today I discovered this "login -> sleep -> resume" method, and tried:
wine explorer /desktop=genshin,3840x2160 YuanShen.exein shellfgto resume the gameI tried another time. This time,
A minor difference: after resume the game process, there's only network re-connection, while game re-loading did not occurred. Then I can play on with FPS unlocked
我复刻出一个比较容易操作的方法： 1、先进游戏，登录进去，等报错一次； 2、回到主界面，登录进游戏，到可以正常游戏的时候切到kde自带的系统监视器：plasma-systemmonitor（我绑定了ctrl+esc快捷键），在进程列表中找到YuanShen.exe，右键使其待机一两分钟，再继续。 3、切回游戏会提示断网重连，之后再无报错。
我没有测试gnome或其他桌面/窗口管理器是否有类似的工具，也不知道plasma-systemmonitor是否能在其他桌面/窗口管理器上使用。
PS：我使用了mkrsym1开发的fpsunlock，没有任何问题。
追记：复制以下内容，粘贴至一个文本文件，另存为xxx.sh，绑定【sh xxx.sh（完整路径）】至快捷键，可以方便地进行操作。 注意：sleep 60可以改为你可以接受的数值，在有些机器上可能60秒不足以使游戏进行断网重连，此时需要改为更可靠的数值。
@kamikaze211 感觉原理和我复刻出来的一样，都是将游戏进程挂起。我没有 “先进游戏，登录进去，等报错一次” 这样的操作，我猜也许你也可以省略这一步？
另外我猜测更核心的是断网重连这一机制。原神 bug 千千万，断网重连占一半。
@littlewu2508 感谢，也许确实可以省略这一步，如果我的手能快一点的话。 原神 bug 千千万，断网重连占一半（还有一半在修BUG，草）
It seems that network re-connection is the way to escape the "Error 10351-4001". After entering the game, either disconnect the network or hang the game process can trigger a re-connection (no need to wait), and then I can play on.
No need to do anything now. Even without disconnecting and reconnecting the network, no more any error report now.
0.0.0.0 public-data-api.mihoyo.com 0.0.0.0 log-upload.mihoyo.com 0.0.0.0 uspider.yuanshen.com
将上面的hosts配置写入/etc/hosts，然后刷新Hosts，理论上可以达到同样的效果。 目前我已经待机十分钟没有任何弹窗。 此方法是参考原来的国服补丁patch.sh脚本，其中一个函数即配置hosts。
@WuJiaQuan 从昨日起已经不再弹窗了，目测是windows用户也有诸多4001弹窗困扰，因此反馈修复了。 顺便，hosts是早就改过的，之前一样弹窗。
@kamikaze211 我刚刚把host恢复后，我还是会弹，但是我把host再配置上去就不回，我系统是ubuntu 22.04 lts
@WuJiaQuan 我的hosts里面有那几行，是很早之前打patch加进去的，不操作挂起的那一步就依然会弹窗，我是用Steam玩的
@codychan 你的情况和我挺像的，我也是用steam玩的，hosts也是很早以前打patch加进去的一直没动（刚才试了下删掉hosts但还是弹），现在的情况同样也是仍然一直弹窗只能用断网大法。
看了你的评论我有点怀疑会不会是在steam下运行的问题，有空我试试脱离steam运行看看还会不会弹。
我是一直有将如下内容加入hosts的（自我开始在linux上玩原神/崩铁以来，少说也有一年多了吧）：
我也是steam启动原神的，用的ge-proton-8-25，不过我是archlinux。
说起近期的4001弹窗，其实一开始是只会弹一次，重新进门后就不会弹了，所以大家也就相安无事。 直到开始无限弹窗，我最初删除原有prefix，也是仅有一次弹窗，好用了几天，然后又无限弹窗了。 再后来有了断网大法。 再到昨天，就不需要断网，也没有弹窗了。
PS：近期在米游社和B站也能看到一些windows用户反馈此问题。
10351-4001 happens again on . I experienced it from 2024-10-13. Re-entering, or
kill -STOP; kill -CONTreconnecting does not help. Using/etc/hostswithdoes not help either.
According to discussions at https://github.com/yaagl/yet-another-anime-game-launcher/issues/466#issuecomment-2420308593,
-platform_type CLOUD_THIRD_PARTY_PCcan resolve the problem.@littlewu2508 Thank you for the cross-reference. I did document the workaround as of 506b77f. Please let me know when the situation changes, or another workaround is needed.