search

bperez77 / xilinx_axidma

A zero-copy Linux driver and a userspace interface library for Xilinx's AXI DMA and VDMA IP blocks. These serve as bridges for communication between the processing system and FPGA programmable logic fabric, through one of the DMA ports on the Zynq processing system. Distributed under the MIT License.
MIT License
472 stars 230 forks source link

can axidma_display_image work on vdma ? #122

Open zskdan opened 4 years ago

zskdan commented 4 years ago

I'm trying to use axidma_display_image to display an image (bitmap image 640 x 480 x 32) on the DISPLAY attached on HDMI to my ARTY-Z7-20. The FPGA base design is the one in https://github.com/Digilent/Arty-Z7-20-base-linux. I've just replaced AXIDMA_DMA with AXIDMA_VDMA references as in this patch 0001-display_image-use-vdma.txt

But It seems that the 2nd copy_from_user() here is making a kernel oops:

Unhandled fault: page domain fault (0x01b) at 0xbe880b58
pgd = dc26c000
[be880b58] *pgd=1d8e6831, *pte=0107b75f, *ppte=0107bc7f
Internal error: Oops - BUG: 1b [#1] PREEMPT SMP ARM
Modules linked in: xilinx_axidma(O) uio_pdrv_genirq
CPU: 1 PID: 1241 Comm: axidma_display_ Tainted: G           O    4.9.0-xilinx-v2017.4 #13
Hardware name: Xilinx Zynq Platform
task: ddb93200 task.stack: dda3a000
PC is at axidma_ioctl+0x7c8/0xb68 [xilinx_axidma]
LR is at axidma_ioctl+0xa88/0xb68 [xilinx_axidma]
pc : [<bf004cc8>]    lr : [<bf004f88>]    psr: a0070013
sp : dda3beb0  ip : 00000007  fp : be880c6c
r10: 00000000  r9 : dda3a000  r8 : dda1b440
r7 : dda3a000  r6 : 00000004  r5 : c0c44c00  r4 : be880b50
r3 : dda3beb0  r2 : 00000000  r1 : 00000007  r0 : 00000007
Flags: NzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
Control: 18c5387d  Table: 1c26c04a  DAC: 00000051
Process axidma_display_ (pid: 1241, stack limit = 0xdda3a210)
Stack: (0xdda3beb0 to 0xdda3c000)
bea0:                                     00000000 00000000 ddba1808 c01a7820
bec0: dc2bf240 00000000 00000001 dda1b440 000001e0 00000280 00000003 de033280
bee0: 00000046 00000000 dc2b1180 dda3bf88 000e1000 dda3bf88 be880b50 dc13ede8
bf00: dc2b1000 80185708 be880b50 c01dcc84 be880b50 c01dce28 000e1000 00000000
bf20: 00000000 00000000 dda3bf1c 00000000 dc2b1180 00000000 00000000 00000000
bf40: 00000000 dc2b1188 00000000 000e1000 ddba1808 00000001 dda3a000 00000000
bf60: be880c6c c01cdbbc 00000000 00000004 dc2b1000 dc2b1000 80185708 be880b50
bf80: dda3a000 00000000 be880c6c c01dd670 00011908 00000000 000109a4 00000036
bfa0: c0106e24 c0106c60 00011908 00000000 00000004 80185708 be880b50 be880b50
bfc0: 00011908 00000000 000109a4 00000036 00000000 00000000 b6f8e000 be880c6c
bfe0: 00022128 be880b24 000114a0 b6ed7e8c 20070010 00000004 00000000 00000000
[<bf004cc8>] (axidma_ioctl [xilinx_axidma]) from [<c01dcc84>] (vfs_ioctl+0x28/0x3c)
[<c01dcc84>] (vfs_ioctl) from [<c01dce28>] (do_vfs_ioctl+0x84/0x898)
[<c01dce28>] (do_vfs_ioctl) from [<c01dd670>] (SyS_ioctl+0x34/0x58)
[<c01dd670>] (SyS_ioctl) from [<c0106c60>] (ret_fast_syscall+0x0/0x3c)
Code: e3a01018 eafffead e1a0300d e0077003 (e5941008) 
---[ end trace bf7c56d3b22f2c8d ]---

Any idea how to resolve that?

dhzzy88 commented 4 years ago

I have the problem,too.

ajaixavier commented 4 years ago

This is a 'page domain fault' crash from kernel. To avoid this, check the status of 'CONFIG_CPU_SW_DOMAIN_PAN' flag in your kernel config. $ zcat /proc/config.gz | grep CONFIG_CPU_SW_DOMAIN_PAN If it is in enabled state, then disable it to avoid 'page domain fault' crash from kernel.

zskdan commented 3 years ago

Thanks @ajaixavier for your feedback. I've just finally got time to go back to this project.

I have no more kernel panic after unsetting CONFIG_CPU_SW_DOMAIN_PAN.

axidma_display_image seems to be running OK. but unfortunately nothing show off from the HDMI. Do you have any working setup for this ?

Thanks!

Yuxin-Yu commented 3 months ago

Hello @zskdan @dhzzy88 @ajaixavier , I encountered an error at the same location, but the error message is different. What could be the reason for this?

[ 1299.997142] Unable to handle kernel access to user memory without uaccess routines at virtual address 0000003fe89688a0
[ 1300.007734] Oops [#1]
[ 1300.009180] Modules linked in: axidma(O)
[ 1300.013100] CPU: 0 PID: 809 Comm: axidma_display_ Tainted: G           O       6.3.12-dirty #6
[ 1300.021664] Hardware name: freechips,rocketchip-vivado (DT)
[ 1300.027222] epc : axidma_ioctl+0x8e8/0xa26 [axidma]
[ 1300.039412]  ra : axidma_ioctl+0x8e8/0xa26 [axidma]
[ 1300.051540] epc : ffffffff01334cba ra : ffffffff01334cba sp : ffffffd8844ffe00
[ 1300.058700]  gp : ffffffff812e3868 tp : ffffffd88747f480 t0 : 6100000000000000
[ 1300.065900]  t1 : 0000000000000061 t2 : 61203a616d646978 s0 : ffffffd8844ffea0
[ 1300.073100]  s1 : 0000003fe8968898 a0 : 0000000000000036 a1 : ffffffd8fc5e0588
[ 1300.080306]  a2 : ffffffd8fc5e3620 a3 : 0000000000000000 a4 : d1be0e6e08481a00
[ 1300.087510]  a5 : d1be0e6e08481a00 a6 : c0000000ffffefff a7 : ffffffff81319700
[ 1300.094722]  s2 : ffffffd8844ffe10 s3 : ffffffff812e40a0 s4 : 0000000000000008
[ 1300.101920]  s5 : ffffffd883869400 s6 : 0000002ad3a36980 s7 : 0000003f8c1dfd10
[ 1300.109128]  s8 : 0000003f8c1e0008 s9 : 0000000000000000 s10: 0000002ad3a03a44
[ 1300.116334]  s11: 0000002ad3a039b8 t3 : ffffffff812f589f t4 : ffffffff812f589f
[ 1300.123536]  t5 : ffffffff812f58a0 t6 : ffffffd8844ffc28
[ 1300.128828] status: 0000000200000120 badaddr: 0000003fe89688a0 cause: 000000000000000d
[ 1300.136766] [<ffffffff01334cba>] axidma_ioctl+0x8e8/0xa26 [axidma]
[ 1300.150258] [<ffffffff80181b08>] sys_ioctl+0x6e/0x8c
[ 1300.155172] [<ffffffff80003570>] ret_from_syscall+0x0/0x2
[ 1300.160760] Code: 0613 6066 3517 0000 0513 3c65 9097 7f38 80e7 ab00 (6488) 85d2 
[ 1300.169406] ---[ end trace 0000000000000000 ]---
Segmentation fault