5.4-rc1: series of Oops in kernel related to lima
Hi, I gave run for vanilla 5.4-rc1 on s905 (mali450) and received series of Ooops in kernel. 5.3 works perfectly in the same environment. Oops is like this:
------------[ cut here ]------------
[ 296.419082] Ignoring spurious kernel translation fault at virtual address ffff800012978020
[ 296.419091] WARNING: CPU: 0 PID: 5797 at arch/arm64/mm/fault.c:302 __do_kernel_fault+0xf8/0x128
[ 296.419093] Modules linked in: zram rpcsec_gss_krb5 ir_rc6_decoder ir_nec_decoder rc_rc6_mce crct10dif_ce lima mceusb gpu_sched meson_vdec(C) rc_tanix_tx3mini meson_ir v4l2_mem2mem rc_core videobuf2_v4l2 meson_rng videobuf2_dma_contig ao_cec videobuf2_memops dw_hdmi_cec videobuf2_common meson_gxbb_wdt pwm_meson videodev mc scpi_hwmon nvmem_meson_efuse uinput ppdev parport loop ipv6 nf_defrag_ipv6 overlay
[ 296.419137] CPU: 0 PID: 5797 Comm: X Tainted: G WC 5.4.0-rc1 #1
[ 296.419140] Hardware name: Oranth Tanix TX3 Mini (DT)
[ 296.419144] pstate: 80000005 (Nzcv daif -PAN -UAO)
[ 296.419149] pc : __do_kernel_fault+0xf8/0x128
[ 296.419155] lr : __do_kernel_fault+0xf8/0x128
[ 296.419157] sp : ffff800012d3b840
[ 296.419160] x29: ffff800012d3b840 x28: ffff0000652e4600
[ 296.419166] x27: 0000000000000000 x26: ffff00005d858080
[ 296.419171] x25: ffff800011379000 x24: 0000000000000025
[ 296.419176] x23: 0000000060000005 x22: ffff800012978020
[ 296.419182] x21: ffff800012d3b8d0 x20: 0000000096000007
[ 296.419187] x19: 0000000000000025 x18: ffffffffffffffff
[ 296.419192] x17: 0000000000000000 x16: 0000000000000000
[ 296.419197] x15: ffff800011379808 x14: 3231303030386666
[ 296.419202] x13: 6666207373657264 x12: 6461206c61757472
[ 296.419208] x11: 697620746120746c x10: 0000000000000000
[ 296.419213] x9 : 0000000000000003 x8 : 0000000000002b46
[ 296.419218] x7 : 0000000000000000 x6 : 0000000009295fe5
[ 296.419223] x5 : 0000000000000003 x4 : 0000000000000001
[ 296.419228] x3 : 0000000000000007 x2 : 0000000000000006
[ 296.419233] x1 : dd487ac8d94f2f00 x0 : 0000000000000000
[ 296.419237] Call trace:
[ 296.419243] __do_kernel_fault+0xf8/0x128
[ 296.419248] do_translation_fault+0x58/0xb0
[ 296.419252] do_mem_abort+0x50/0xb0
[ 296.419256] el1_da+0x20/0x94
[ 296.419261] drm_gem_fence_array_add+0x60/0x198
[ 296.419266] drm_gem_fence_array_add_implicit+0x94/0x1a8
[ 296.419275] lima_gem_submit+0x464/0x518 [lima]
[ 296.419284] lima_ioctl_gem_submit+0x270/0x2d0 [lima]
[ 296.419289] drm_ioctl_kernel+0xd0/0x128
[ 296.419293] drm_ioctl+0x1fc/0x430
[ 296.419298] do_vfs_ioctl+0xc4/0x928
[ 296.419303] ksys_ioctl+0x50/0x80
[ 296.419308] __arm64_sys_ioctl+0x28/0x38
[ 296.419314] el0_svc_handler+0x134/0x228
[ 296.419318] el0_svc+0x10/0x14
[ 296.419322] ---[ end trace 9b9339ca351c6eaf ]---
looking on dmesg I see suspicious entries regarding lima:
[ 7.406630] lima d00c0000.gpu: IRQ ppmmu3 not found
[ 7.406645] lima d00c0000.gpu: IRQ ppmmu4 not found
[ 7.406652] lima d00c0000.gpu: IRQ ppmmu5 not found
[ 7.406660] lima d00c0000.gpu: IRQ ppmmu6 not found
[ 7.406667] lima d00c0000.gpu: IRQ ppmmu7 not found
[ 7.406802] lima d00c0000.gpu: gp - mali450 version major 0 minor 0
[ 7.407092] lima d00c0000.gpu: pp0 - mali450 version major 0 minor 0
[ 7.407323] lima d00c0000.gpu: pp1 - mali450 version major 0 minor 0
[ 7.407398] lima d00c0000.gpu: pp2 - mali450 version major 0 minor 0
[ 7.407447] lima d00c0000.gpu: IRQ pp3 not found
[ 7.407455] lima d00c0000.gpu: IRQ pp4 not found
[ 7.407462] lima d00c0000.gpu: IRQ pp5 not found
[ 7.407469] lima d00c0000.gpu: IRQ pp6 not found
[ 7.407476] lima d00c0000.gpu: IRQ pp7 not found
[ 7.407486] lima d00c0000.gpu: l2 cache 8K, 4-way, 64byte cache line, 128bit external bus
[ 7.407495] lima d00c0000.gpu: l2 cache 64K, 4-way, 64byte cache line, 128bit external bus
[ 7.420936] lima d00c0000.gpu: bus rate = 166666667
[ 7.420950] lima d00c0000.gpu: mod rate = 744000000
[ 7.428306] [drm] Initialized lima 1.0.0 20190217 for d00c0000.gpu on minor 1