Hi,
Recently, when test IPCF feature, we found that, the ipc-shm-dev interrupt are still raising, after run the ipcf sample demo(single instance), and the number of interrupts are in rapid growth, like the attached log(ipcf-interrupt-issue-based-on-SDK-BSP37&38-on-RDB3.log) shows.
The issue comes out after SDK BSP37&38 releases, moreover, the related commits seems the following:
commit 2c269b3
Author: Hoang Nguyen Le <hoang.nguyenle@nxp.com>
Date: Thu Mar 16 06:44:45 2023 +0100
Correct irq configuration for hw of s32g3
- Restructure hw directory
- Update sample with specific configuration files
Issue: IPCF-2945
Signed-off-by: Hoang Nguyen Le <hoang.nguyenle@nxp.com>
Would you please help to check the issue and give a fix or workaround?
Thanks,
Zhantao
Hi,
Can you let us know which IPCF version are you using with BSP37.0/38.0?
Please, let us know.
Hi,
Thanks for the notice. We are not seeing anything on regards of this behavior. Let us investigate and understand if there is an available workaround for this.
Just to understand from our side, is this happening on both S32G2 and S32G3? Or just under S32G3 chip?
Please, let us know.
We mainly tested the S32G3 board(RDB3), and I will test S32G2 board to check the status, but
I think there maybe same issue on S32G2, anyway, I will give the S32G2 result tomorrow.
Thanks,
Zhantao
Hi,
Thanks for your feedback. It has been confirmed to be a bug and the related team has been noticed.
We do apologize for the inconvenience.
Please, let us know.
Hi @Daniel-Aguirre,
We notice that, there is a new v4.9 IPCF firmware released, and it helps to fix the interrupt issue. I tested IPCF feature with this v4.9 firmwares for both S32G2 and S32G3 boards, the interrupt counts of ipc-shm device are correct with these new firmwares, but there is a new issue, that the "ksoftirqd/0" will occupy 100.0 %CPU in "top".
It only happens when ipc-shm kos insmoded and the ping demos executed, if rmmod the ipc-shm kos, the ksoftirqd issue will disappear.
Would you please help to confirm the issue and find the root reason?
Thanks,
Zhantao
Hi,
We do apologize for the delay. We have received the following update from the internal team:
"This issue was reported near the release schedule of the last version of IPCF (IPCF 4.9.0). So there was no time to fix this issue on IPCF4.9.0."
We do apologize.
Please, let us know.
Hi,
Thanks for your update. Let us verify this issue.
Once we have an update, we will let you know. We apologize for any delay from our side.
Please, let us know.