-
Notifications
You must be signed in to change notification settings - Fork 35
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Facing kernel panic for openEuler 23.09 #100
Comments
Hi Sadiq, thanks for your comments! @fengerhu1 @Fly0307 Erhu and Xi, please help check the issue and help Sadiq to fix it out. Thanks! |
Hi Sadiq, We're sorry you're having this problem with penglai. |
Add a sbi call to reclaim memory and clear pmp settings when uninstalling the pontoon module; add the release operation for kmalloc data Penglai-Enclave#100
This is an issue we have fixed in the latest version. |
Hi Dong Du, Fly0307.
As per the above log should I use multilib toolchain?
Please let me know your suggestions or comments. Thanks, |
For sdk compiling, we've provided a docker image that contains most of the toolchain for penglai, the toolchain doesn't need to support multilib. you can start the image with this command:
You can compile the sdk by executing the following command in the penglai-enclave-sPMP directory:
|
Hi, Please let me know your comments. Thanks, |
@Sadiq-techie Thanks, |
Hi Erhu Feng, I want to use the same version of cross-compilier to build sdk, driver and opensbi. Thanks, |
Hi Erhu Feng,
In 1 out of 4 iterations observed this issue, driver cleanup is not occurring correctly during the module unloading process as below
Please let me know your comments. Thanks, |
@Sadiq-techie Can you provide more information about the execution instructions and the output before and after? |
We have fixed this issue in the latest version, you can try it at the following commit |
Hi Erhu Feng, When I was attempting to verify the TEE with the latest sources I was facing below issue
Please let me know your comments. Thanks, |
Hi Erhu Feng, After the above mentioned issue, now unable to boot the kernel and stuck with below issue
Please let me know your comments. Thanks, |
You can fix this with the following patch:0001-fix-2-power-problem.patch |
Thanks Erhu Feng for the quick response, will apply this patch and verify. |
Hi Erhu Feng, When I was attempting to verify the TEE with the latest sources I was facing below issue
Please let me know your comments or suggestions. Thanks, |
Hi Erhu Feng, below are the observations with the latest sources observed the kernel bug. with this commit 9a5e2dc there is no kernel bug, demo prime is working as expected.
After uninstalling the penglai module correctly it output the following message:
Instead of below message Final output:
` Thanks, |
You're right, when using the latest commit test and closing the debug output, it prints log like this:
|
This message does not affect the normal execution of the program, if necessary you can solve this problem with this patch |
Hi Dong Du,
Facing kernel panic for openEuler 23.09
Workstation details:
Ubuntu 22.04
RISC-V cross compilier 13.2.0
qemu version 8.1.50
Steps to reproduce:
Cross compilied u-boot, opensbi-1.2 as per the steps for openEuler >23.
when cross compiling sdk it was unsuccessful.
used the https://ipads.se.sjtu.edu.cn:1313/d/6a464e02cd3d4c1bafb0/ sdk source and cross compilation was successful.
Booted the openEuler 23.09 with below command
copied files from host to openEuler qemu as follows:
cross compiled penglai-kernel-driver generated penglai.ko
insert the enclave driver module
insmod penglai.ko
prime enclave execution was successful
./host prime
there is kernel call trace after enclave execution and kernel panic after some seconds
Similarly for 23.03 followed same steps with replacing openeuler 23.09 with 23.03 and was facing kernel panic as follows:
openEuler 23.03 uses 6.1.x Linux kernel and openEuler 23.09 uses 6.4.x Linux kernel.
observations:
when booted with openEuler 23.03 it is observed mmu was sv57 and with openEuler 23.09 it was sv39.
Any issue with OpenSBI?
Please let me know your suggestions or comments.
Thanks,
MD Sadiq
The text was updated successfully, but these errors were encountered: