mbox

[PULL,00/14] MIPS patches for 2024-10-31

Message ID 20241031042130.98450-1-philmd@linaro.org
State New
Headers show

Pull-request

https://github.com/philmd/qemu.git tags/mips-20241031

Message

Philippe Mathieu-Daudé Oct. 31, 2024, 4:21 a.m. UTC 
The following changes since commit 58d49b5895f2e0b5cfe4b2901bf24f3320b74f29:

  Merge tag 'net-pull-request' of https://github.com/jasowang/qemu into staging (2024-10-29 14:00:43 +0000)

are available in the Git repository at:

  https://github.com/philmd/qemu.git tags/mips-20241031

for you to fetch changes up to dcc3c68c844fef7bc018ab53f0993b8d44137d77:

  target/mips: Remove unused CPUMIPSState::current_fpu field (2024-10-31 00:48:45 -0300)

----------------------------------------------------------------
MIPS patches queue

- Migrate missing CP0 TLB MemoryMapID register (Yongbok)
- Enable MSA ASE for mips32r6-generic (Aleksandar)
- Convert Loongson LEXT opcodes to decodetree (Phil)
- Introduce ase_3d_available and disas_mt_available helpers (Phil)

----------------------------------------------------------------

Aleksandar Markovic (1):
  target/mips: Enable MSA ASE for mips32r6-generic

Philippe Mathieu-Daudé (12):
  target/mips: Extract decode_64bit_enabled() helper
  target/mips: Simplify Loongson MULTU.G opcode
  target/mips: Re-introduce OPC_ADDUH_QB_DSP and OPC_MUL_PH_DSP
  target/mips: Convert Loongson DDIV.G opcodes to decodetree
  target/mips: Convert Loongson DIV.G opcodes to decodetree
  target/mips: Convert Loongson [D]DIVU.G opcodes to decodetree
  target/mips: Convert Loongson [D]MOD[U].G opcodes to decodetree
  target/mips: Convert Loongson [D]MULT[U].G opcodes to decodetree
  target/mips: Remove unreachable 32-bit code on 64-bit Loongson Ext
  target/mips: Introduce ase_3d_available() helper
  target/mips: Introduce disas_mt_available()
  target/mips: Remove unused CPUMIPSState::current_fpu field

Yongbok Kim (1):
  target/mips: Migrate TLB MemoryMapID register

 target/mips/cpu.h                         |   7 +-
 target/mips/mips-defs.h                   |   2 -
 target/mips/tcg/translate.h               |   8 +
 target/mips/tcg/godson2.decode            |  27 ++
 target/mips/tcg/loong-ext.decode          |  28 ++
 target/mips/sysemu/machine.c              |   9 +-
 target/mips/tcg/loong_translate.c         | 271 ++++++++++++++
 target/mips/tcg/translate.c               | 415 +++++-----------------
 target/mips/cpu-defs.c.inc                |  16 +-
 target/mips/tcg/micromips_translate.c.inc |   5 +-
 target/mips/tcg/meson.build               |   3 +
 11 files changed, 452 insertions(+), 339 deletions(-)
 create mode 100644 target/mips/tcg/godson2.decode
 create mode 100644 target/mips/tcg/loong-ext.decode
 create mode 100644 target/mips/tcg/loong_translate.c

Comments

Peter Maydell Nov. 1, 2024, 4:44 p.m. UTC | #1 
On Thu, 31 Oct 2024 at 04:22, Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
>
> The following changes since commit 58d49b5895f2e0b5cfe4b2901bf24f3320b74f29:
>
>   Merge tag 'net-pull-request' of https://github.com/jasowang/qemu into staging (2024-10-29 14:00:43 +0000)
>
> are available in the Git repository at:
>
>   https://github.com/philmd/qemu.git tags/mips-20241031
>
> for you to fetch changes up to dcc3c68c844fef7bc018ab53f0993b8d44137d77:
>
>   target/mips: Remove unused CPUMIPSState::current_fpu field (2024-10-31 00:48:45 -0300)
>
> ----------------------------------------------------------------
> MIPS patches queue
>
> - Migrate missing CP0 TLB MemoryMapID register (Yongbok)
> - Enable MSA ASE for mips32r6-generic (Aleksandar)
> - Convert Loongson LEXT opcodes to decodetree (Phil)
> - Introduce ase_3d_available and disas_mt_available helpers (Phil)
>
> ----------------------------------------------------------------

Hi; it looks like this breaks the foolong2e functional test:

https://gitlab.com/qemu-project/qemu/-/jobs/8249048723

12/15 qemu:func-thorough+func-mips64el-thorough+thorough /
func-mips64el-mips64el_fuloong2e ERROR 2.54s

Guest kernel panics rather than booting.

thanks
-- PMM
Philippe Mathieu-Daudé Nov. 4, 2024, 10:46 a.m. UTC | #2 
Hi,

TL;DR: I will repost the PR will the typo fixed.

On 1/11/24 13:44, Peter Maydell wrote:
> On Thu, 31 Oct 2024 at 04:22, Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
>>
>> The following changes since commit 58d49b5895f2e0b5cfe4b2901bf24f3320b74f29:
>>
>>    Merge tag 'net-pull-request' of https://github.com/jasowang/qemu into staging (2024-10-29 14:00:43 +0000)
>>
>> are available in the Git repository at:
>>
>>    https://github.com/philmd/qemu.git tags/mips-20241031
>>
>> for you to fetch changes up to dcc3c68c844fef7bc018ab53f0993b8d44137d77:
>>
>>    target/mips: Remove unused CPUMIPSState::current_fpu field (2024-10-31 00:48:45 -0300)
>>
>> ----------------------------------------------------------------
>> MIPS patches queue
>>
>> - Migrate missing CP0 TLB MemoryMapID register (Yongbok)
>> - Enable MSA ASE for mips32r6-generic (Aleksandar)
>> - Convert Loongson LEXT opcodes to decodetree (Phil)
>> - Introduce ase_3d_available and disas_mt_available helpers (Phil)
>>
>> ----------------------------------------------------------------
> 
> Hi; it looks like this breaks the foolong2e functional test:
> 
> https://gitlab.com/qemu-project/qemu/-/jobs/8249048723
> 
> 12/15 qemu:func-thorough+func-mips64el-thorough+thorough /
> func-mips64el-mips64el_fuloong2e ERROR 2.54s
> 
> Guest kernel panics rather than booting.

Oops. At least we have a test. I suppose I missed it because the job
isn't run by default and has been renamed due to the avocado ->
functional conversion.

The bug is a due to a typo:

-- >8 --
--- a/target/mips/tcg/loong_translate.c
+++ b/target/mips/tcg/loong_translate.c
@@ -68,7 +68,7 @@ static bool gen_lext_DIV_G(DisasContext *s, int rd, 
int rs, int rt)

  static bool trans_DDIV_G(DisasContext *s, arg_muldiv *a)
  {
-    return gen_lext_DIV_G(s, a->rt, a->rs, a->rd);
+    return gen_lext_DIV_G(s, a->rd, a->rs, a->rt);
  }
---

But then the test fail later:

$ ninja qemu-system-mips64el && ./qemu-system-mips64el -kernel 
boot/vmlinux-3.16.0-6-loongson-2e -machine fuloong2e -append 
'console=ttyS0' -serial mon:stdio -d int -trace pci\*
...
[    0.000000] Kernel command line: console=ttyS0
[    0.436000] PCI host bridge to bus 0000:00
[    0.436000] pci_bus 0000:00: root bus resource [mem 
0x14000000-0x1c000000]
[    0.440000] pci_bus 0000:00: root bus resource [io  0x4000-0xffff]
[    0.440000] pci_bus 0000:00: No busn resource found for root bus, 
will use [bus 00-ff]
...
pci_cfg_read ati-vga 00:06.0 @0x0 -> 0x51591002
pci_cfg_read ati-vga 00:06.0 @0xc -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x8 -> 0x3000000
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x10 -> 0x8
pci_cfg_write ati-vga 00:06.0 @0x10 <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x10 -> 0xff000008
pci_cfg_write ati-vga 00:06.0 @0x10 <- 0x8
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x14 -> 0x1
pci_cfg_write ati-vga 00:06.0 @0x14 <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x14 -> 0xffffff01
pci_cfg_write ati-vga 00:06.0 @0x14 <- 0x1
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x18 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x18 <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x18 -> 0xffffc000
pci_cfg_write ati-vga 00:06.0 @0x18 <- 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x1c -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x1c <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x1c -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x1c <- 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x20 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x20 <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x20 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x20 <- 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x24 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x24 <- 0xffffffff
pci_cfg_read ati-vga 00:06.0 @0x24 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x24 <- 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x30 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x30 <- 0xfffff800
pci_cfg_read ati-vga 00:06.0 @0x30 -> 0xffff0000
pci_cfg_write ati-vga 00:06.0 @0x30 <- 0x0
pci_cfg_read ati-vga 00:06.0 @0x2c -> 0x11001af4
pci_cfg_read ati-vga 00:06.0 @0x2c -> 0x11001af4
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
[    0.464000] vgaarb: device added: 
PCI:0000:00:06.0,decodes=io+mem,owns=none,locks=none
...
[    0.472000] pci 0000:00:06.0: BAR 0: assigned [mem 
0x14000000-0x14ffffff pref]
pci_cfg_write ati-vga 00:06.0 @0x10 <- 0x14000008
pci_cfg_read ati-vga 00:06.0 @0x10 -> 0x14000008
[    0.472000] pci 0000:00:07.0: BAR 6: assigned [mem 
0x15000000-0x1503ffff pref]
[    0.472000] pci 0000:00:06.0: BAR 6: assigned [mem 
0x15040000-0x1504ffff pref]
[    0.472000] pci 0000:00:06.0: BAR 2: assigned [mem 0x15050000-0x15053fff]
pci_cfg_write ati-vga 00:06.0 @0x18 <- 0x15050000
pci_cfg_read ati-vga 00:06.0 @0x18 -> 0x15050000
[    0.472000] pci 0000:00:05.5: BAR 0: assigned [io  0x4000-0x40ff]
...
[    0.476000] pci 0000:00:06.0: BAR 1: assigned [io  0x4400-0x44ff]
pci_cfg_write ati-vga 00:06.0 @0x14 <- 0x4401
pci_cfg_read ati-vga 00:06.0 @0x14 -> 0x4401
...
pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
pci_cfg_write ati-vga 00:06.0 @0x3c <- 0x13a
...
pci_cfg_read ati-vga 00:06.0 @0xc -> 0x0
...
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
[    0.864000] PCI: Enabling device 0000:00:06.0 (0000 -> 0003)
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
pci_cfg_write ati-vga 00:06.0 @0x4 <- 0x3
pci_update_mappings_add ati-vga 00:06.0 0,0x14000000+0x1000000
pci_update_mappings_add ati-vga 00:06.0 1,0x4400+0x100
pci_update_mappings_add ati-vga 00:06.0 2,0x15050000+0x4000
pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x13a
pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x3
do_raise_exception_err: 28 (data bus error) 0
mips_cpu_do_interrupt enter: PC ffffffff8045aed8 EPC ffffffff8043ae88 
data bus error exception
mips_cpu_do_interrupt: PC ffffffff80000180 EPC ffffffff8045aed8 cause 7
     S 1400a4e3 C 0000001c A c000000000010000 D 0000000000000000
[    0.868000] Data bus error, epc == ffffffff8045aed8, ra == 
ffffffff8045aeb8
[    0.868000] Oops[#1]:
[    0.868000] CPU: 0 PID: 1 Comm: swapper Not tainted 
3.16.0-6-loongson-2e #1 Debian 3.16.56-1+deb8u1
[    0.868000] task: 980000000e06f580 ti: 980000000e078000 task.ti: 
980000000e078000
[    0.868000] $ 0   : 0000000000000000 ffffffff80428db0 
9000000000000000 0000000015050000
[    0.868000] $ 4   : 9000000015050000 ffffffff80430000 
0000000015053fff 0000000000000070
[    0.868000] $ 8   : 0000000000000000 ffffffff8079b190 
0000000000000001 fffffffffffffff8
[    0.868000] $12   : 0000000000000000 ffffffff803fd67c 
0000000000000000 0000000000000082
[    0.868000] $16   : 980000000e278000 980000000e336000 
980000000e336428 0000000000000000
[    0.868000] $20   : ffffffff8087f918 0000000000000000 
ffffffff80929cf0 ffffffff80800000
[    0.868000] $24   : 0000000000000018 0000000000000000 

[    0.868000] $28   : 980000000e078000 980000000e07bbc0 
0000000000000074 ffffffff8045aeb8
[    0.868000] Hi    : 0000000000d2f000
[    0.868000] Lo    : 0000000000fa7d00
[    0.868000] epc   : ffffffff8045aed8 radeonfb_pci_register+0x1b0/0x1ad0
[    0.868000]     Not tainted
[    0.868000] ra    : ffffffff8045aeb8 radeonfb_pci_register+0x190/0x1ad0
[    0.868000] Status: 1400a4e3	KX SX UX KERNEL EXL IE
[    0.868000] Cause : 0000001c
[    0.868000] PrId  : 00006302 (ICT Loongson-2)
[    0.868000] Modules linked in:
[    0.868000] Process swapper (pid: 1, threadinfo=980000000e078000, 
task=980000000e06f580, tls=0000000000000000)
[    0.868000] Stack : 980000000e336748 980000000e278090 
000000001400a4e1 ffffffff8087f670
	  ffffffff8087f918 0000000000000000 980000000e278090 ffffffff8087f6d8
	  980000000e278000 ffffffff8087f670 ffffffff8087f918 0000000000000000
	  ffffffff80929cf0 ffffffff80800000 0000000000000074 ffffffff8042e718
	  980000000e278090 ffffffff809b0000 980000000e2780f0 ffffffff8087f6d8
	  ffffffff809b0000 ffffffff804a3dcc 980000000e355380 980000000e278090
	  ffffffff8087f6d8 980000000e2780f0 0000000000000000 ffffffff80810000
	  0000000000000000 ffffffff804a4238 0000000000000000 ffffffff8087f6d8
	  ffffffff804a4168 ffffffff804a1a04 980000000e1c6090 980000000e1cbde0
	  ffffffff8087f6d8 980000000e355380 ffffffff8087e8f0 ffffffff804a3218
	  ...
[    0.868000] Call Trace:
[    0.868000] [<ffffffff8045aed8>] radeonfb_pci_register+0x1b0/0x1ad0
[    0.868000] [<ffffffff8042e718>] pci_device_probe+0xb8/0x130
[    0.868000] [<ffffffff804a3dcc>] driver_probe_device+0xdc/0x400
[    0.868000] [<ffffffff804a4238>] __driver_attach+0xd0/0xd8
[    0.868000] [<ffffffff804a1a04>] bus_for_each_dev+0x7c/0xc8
[    0.868000] [<ffffffff804a3218>] bus_add_driver+0x1a0/0x288
[    0.868000] [<ffffffff804a4c10>] driver_register+0x90/0x138
[    0.868000] [<ffffffff808e2444>] radeonfb_init+0x60/0x230
[    0.868000] [<ffffffff801005c0>] do_one_initcall+0x110/0x220
[    0.868000] [<ffffffff808b8e3c>] kernel_init_freeable+0x194/0x250
[    0.868000] [<ffffffff80661f98>] kernel_init+0x20/0x118
[    0.868000] [<ffffffff80103ed0>] ret_from_kernel_thread+0x14/0x1c
[    0.868000]
[    0.868000]
Code: 0082202d  108005ae  fe440368 <8c850148> 924207d8  00052c00 
0005283c  0005283e  fe450378
[    0.888000] ---[ end trace 2fdf3be6a36f9b46 ]---
[    0.892000] Kernel panic - not syncing: Attempted to kill init! 
exitcode=0x0000000b
[    0.892000]
[    0.892000] ---[ end Kernel panic - not syncing: Attempted to kill 
init! exitcode=0x0000000b
[    0.892000]

I don't recall this failure. Zoltan, does it ring a bell?
Anyway I'll open a gitlab issue to track it, and will repost the
PR will the typo fixed.

Regards,

Phil.
BALATON Zoltan Nov. 4, 2024, 11:32 a.m. UTC | #3 
On Mon, 4 Nov 2024, Philippe Mathieu-Daudé wrote:
> Hi,
>
> TL;DR: I will repost the PR will the typo fixed.
>
> On 1/11/24 13:44, Peter Maydell wrote:
>> On Thu, 31 Oct 2024 at 04:22, Philippe Mathieu-Daudé <philmd@linaro.org> 
>> wrote:
>>> 
>>> The following changes since commit 
>>> 58d49b5895f2e0b5cfe4b2901bf24f3320b74f29:
>>>
>>>    Merge tag 'net-pull-request' of https://github.com/jasowang/qemu into 
>>> staging (2024-10-29 14:00:43 +0000)
>>> 
>>> are available in the Git repository at:
>>>
>>>    https://github.com/philmd/qemu.git tags/mips-20241031
>>> 
>>> for you to fetch changes up to dcc3c68c844fef7bc018ab53f0993b8d44137d77:
>>>
>>>    target/mips: Remove unused CPUMIPSState::current_fpu field (2024-10-31 
>>> 00:48:45 -0300)
>>> 
>>> ----------------------------------------------------------------
>>> MIPS patches queue
>>> 
>>> - Migrate missing CP0 TLB MemoryMapID register (Yongbok)
>>> - Enable MSA ASE for mips32r6-generic (Aleksandar)
>>> - Convert Loongson LEXT opcodes to decodetree (Phil)
>>> - Introduce ase_3d_available and disas_mt_available helpers (Phil)
>>> 
>>> ----------------------------------------------------------------
>> 
>> Hi; it looks like this breaks the foolong2e functional test:
>> 
>> https://gitlab.com/qemu-project/qemu/-/jobs/8249048723
>> 
>> 12/15 qemu:func-thorough+func-mips64el-thorough+thorough /
>> func-mips64el-mips64el_fuloong2e ERROR 2.54s
>> 
>> Guest kernel panics rather than booting.
>
> Oops. At least we have a test. I suppose I missed it because the job
> isn't run by default and has been renamed due to the avocado ->
> functional conversion.
>
> The bug is a due to a typo:
>
> -- >8 --
> --- a/target/mips/tcg/loong_translate.c
> +++ b/target/mips/tcg/loong_translate.c
> @@ -68,7 +68,7 @@ static bool gen_lext_DIV_G(DisasContext *s, int rd, int rs, 
> int rt)
>
> static bool trans_DDIV_G(DisasContext *s, arg_muldiv *a)
> {
> -    return gen_lext_DIV_G(s, a->rt, a->rs, a->rd);
> +    return gen_lext_DIV_G(s, a->rd, a->rs, a->rt);
> }
> ---
>
> But then the test fail later:
>
> $ ninja qemu-system-mips64el && ./qemu-system-mips64el -kernel 
> boot/vmlinux-3.16.0-6-loongson-2e -machine fuloong2e -append 'console=ttyS0' 
> -serial mon:stdio -d int -trace pci\*
> ...
> [    0.000000] Kernel command line: console=ttyS0
> [    0.436000] PCI host bridge to bus 0000:00
> [    0.436000] pci_bus 0000:00: root bus resource [mem 0x14000000-0x1c000000]
> [    0.440000] pci_bus 0000:00: root bus resource [io  0x4000-0xffff]
> [    0.440000] pci_bus 0000:00: No busn resource found for root bus, will use 
> [bus 00-ff]
> ...
> pci_cfg_read ati-vga 00:06.0 @0x0 -> 0x51591002
> pci_cfg_read ati-vga 00:06.0 @0xc -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x8 -> 0x3000000
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
> pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x10 -> 0x8
> pci_cfg_write ati-vga 00:06.0 @0x10 <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x10 -> 0xff000008
> pci_cfg_write ati-vga 00:06.0 @0x10 <- 0x8
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x14 -> 0x1
> pci_cfg_write ati-vga 00:06.0 @0x14 <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x14 -> 0xffffff01
> pci_cfg_write ati-vga 00:06.0 @0x14 <- 0x1
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x18 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x18 <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x18 -> 0xffffc000
> pci_cfg_write ati-vga 00:06.0 @0x18 <- 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x1c -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x1c <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x1c -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x1c <- 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x20 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x20 <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x20 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x20 <- 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x24 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x24 <- 0xffffffff
> pci_cfg_read ati-vga 00:06.0 @0x24 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x24 <- 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x30 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x30 <- 0xfffff800
> pci_cfg_read ati-vga 00:06.0 @0x30 -> 0xffff0000
> pci_cfg_write ati-vga 00:06.0 @0x30 <- 0x0
> pci_cfg_read ati-vga 00:06.0 @0x2c -> 0x11001af4
> pci_cfg_read ati-vga 00:06.0 @0x2c -> 0x11001af4
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> [    0.464000] vgaarb: device added: 
> PCI:0000:00:06.0,decodes=io+mem,owns=none,locks=none
> ...
> [    0.472000] pci 0000:00:06.0: BAR 0: assigned [mem 0x14000000-0x14ffffff 
> pref]
> pci_cfg_write ati-vga 00:06.0 @0x10 <- 0x14000008
> pci_cfg_read ati-vga 00:06.0 @0x10 -> 0x14000008
> [    0.472000] pci 0000:00:07.0: BAR 6: assigned [mem 0x15000000-0x1503ffff 
> pref]
> [    0.472000] pci 0000:00:06.0: BAR 6: assigned [mem 0x15040000-0x1504ffff 
> pref]
> [    0.472000] pci 0000:00:06.0: BAR 2: assigned [mem 0x15050000-0x15053fff]
> pci_cfg_write ati-vga 00:06.0 @0x18 <- 0x15050000
> pci_cfg_read ati-vga 00:06.0 @0x18 -> 0x15050000
> [    0.472000] pci 0000:00:05.5: BAR 0: assigned [io  0x4000-0x40ff]
> ...
> [    0.476000] pci 0000:00:06.0: BAR 1: assigned [io  0x4400-0x44ff]
> pci_cfg_write ati-vga 00:06.0 @0x14 <- 0x4401
> pci_cfg_read ati-vga 00:06.0 @0x14 -> 0x4401
> ...
> pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
> pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x100
> pci_cfg_write ati-vga 00:06.0 @0x3c <- 0x13a
> ...
> pci_cfg_read ati-vga 00:06.0 @0xc -> 0x0
> ...
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> [    0.864000] PCI: Enabling device 0000:00:06.0 (0000 -> 0003)
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x0
> pci_cfg_write ati-vga 00:06.0 @0x4 <- 0x3
> pci_update_mappings_add ati-vga 00:06.0 0,0x14000000+0x1000000
> pci_update_mappings_add ati-vga 00:06.0 1,0x4400+0x100
> pci_update_mappings_add ati-vga 00:06.0 2,0x15050000+0x4000
> pci_cfg_read ati-vga 00:06.0 @0x3c -> 0x13a
> pci_cfg_read ati-vga 00:06.0 @0x4 -> 0x3
> do_raise_exception_err: 28 (data bus error) 0
> mips_cpu_do_interrupt enter: PC ffffffff8045aed8 EPC ffffffff8043ae88 data 
> bus error exception
> mips_cpu_do_interrupt: PC ffffffff80000180 EPC ffffffff8045aed8 cause 7
>    S 1400a4e3 C 0000001c A c000000000010000 D 0000000000000000
> [    0.868000] Data bus error, epc == ffffffff8045aed8, ra == 
> ffffffff8045aeb8
> [    0.868000] Oops[#1]:
> [    0.868000] CPU: 0 PID: 1 Comm: swapper Not tainted 3.16.0-6-loongson-2e 
> #1 Debian 3.16.56-1+deb8u1
> [    0.868000] task: 980000000e06f580 ti: 980000000e078000 task.ti: 
> 980000000e078000
> [    0.868000] $ 0   : 0000000000000000 ffffffff80428db0 9000000000000000 
> 0000000015050000
> [    0.868000] $ 4   : 9000000015050000 ffffffff80430000 0000000015053fff 
> 0000000000000070
> [    0.868000] $ 8   : 0000000000000000 ffffffff8079b190 0000000000000001 
> fffffffffffffff8
> [    0.868000] $12   : 0000000000000000 ffffffff803fd67c 0000000000000000 
> 0000000000000082
> [    0.868000] $16   : 980000000e278000 980000000e336000 980000000e336428 
> 0000000000000000
> [    0.868000] $20   : ffffffff8087f918 0000000000000000 ffffffff80929cf0 
> ffffffff80800000
> [    0.868000] $24   : 0000000000000018 0000000000000000 
> [    0.868000] $28   : 980000000e078000 980000000e07bbc0 0000000000000074 
> ffffffff8045aeb8
> [    0.868000] Hi    : 0000000000d2f000
> [    0.868000] Lo    : 0000000000fa7d00
> [    0.868000] epc   : ffffffff8045aed8 radeonfb_pci_register+0x1b0/0x1ad0
> [    0.868000]     Not tainted
> [    0.868000] ra    : ffffffff8045aeb8 radeonfb_pci_register+0x190/0x1ad0
> [    0.868000] Status: 1400a4e3	KX SX UX KERNEL EXL IE
> [    0.868000] Cause : 0000001c
> [    0.868000] PrId  : 00006302 (ICT Loongson-2)
> [    0.868000] Modules linked in:
> [    0.868000] Process swapper (pid: 1, threadinfo=980000000e078000, 
> task=980000000e06f580, tls=0000000000000000)
> [    0.868000] Stack : 980000000e336748 980000000e278090 000000001400a4e1 
> ffffffff8087f670
> 	  ffffffff8087f918 0000000000000000 980000000e278090 ffffffff8087f6d8
> 	  980000000e278000 ffffffff8087f670 ffffffff8087f918 0000000000000000
> 	  ffffffff80929cf0 ffffffff80800000 0000000000000074 ffffffff8042e718
> 	  980000000e278090 ffffffff809b0000 980000000e2780f0 ffffffff8087f6d8
> 	  ffffffff809b0000 ffffffff804a3dcc 980000000e355380 980000000e278090
> 	  ffffffff8087f6d8 980000000e2780f0 0000000000000000 ffffffff80810000
> 	  0000000000000000 ffffffff804a4238 0000000000000000 ffffffff8087f6d8
> 	  ffffffff804a4168 ffffffff804a1a04 980000000e1c6090 980000000e1cbde0
> 	  ffffffff8087f6d8 980000000e355380 ffffffff8087e8f0 ffffffff804a3218
> 	  ...
> [    0.868000] Call Trace:
> [    0.868000] [<ffffffff8045aed8>] radeonfb_pci_register+0x1b0/0x1ad0
> [    0.868000] [<ffffffff8042e718>] pci_device_probe+0xb8/0x130
> [    0.868000] [<ffffffff804a3dcc>] driver_probe_device+0xdc/0x400
> [    0.868000] [<ffffffff804a4238>] __driver_attach+0xd0/0xd8
> [    0.868000] [<ffffffff804a1a04>] bus_for_each_dev+0x7c/0xc8
> [    0.868000] [<ffffffff804a3218>] bus_add_driver+0x1a0/0x288
> [    0.868000] [<ffffffff804a4c10>] driver_register+0x90/0x138
> [    0.868000] [<ffffffff808e2444>] radeonfb_init+0x60/0x230
> [    0.868000] [<ffffffff801005c0>] do_one_initcall+0x110/0x220
> [    0.868000] [<ffffffff808b8e3c>] kernel_init_freeable+0x194/0x250
> [    0.868000] [<ffffffff80661f98>] kernel_init+0x20/0x118
> [    0.868000] [<ffffffff80103ed0>] ret_from_kernel_thread+0x14/0x1c
> [    0.868000]
> [    0.868000]
> Code: 0082202d  108005ae  fe440368 <8c850148> 924207d8  00052c00 0005283c 
> 0005283e  fe450378
> [    0.888000] ---[ end trace 2fdf3be6a36f9b46 ]---
> [    0.892000] Kernel panic - not syncing: Attempted to kill init! 
> exitcode=0x0000000b
> [    0.892000]
> [    0.892000] ---[ end Kernel panic - not syncing: Attempted to kill init! 
> exitcode=0x0000000b
> [    0.892000]
>
> I don't recall this failure. Zoltan, does it ring a bell?
> Anyway I'll open a gitlab issue to track it, and will repost the
> PR will the typo fixed.

Hello,

No, I haven't seen this before. There were no changes to ati-vga for a 
long time so it's unlikely it was broken by that recently. The ati-vga 
does have a 60Hz vblank interrupt added in b7105d280cab so maybe you could 
try disabling that for testing to see if this has anything to do with the 
above but if it broke recently maybe it's more related to some MIPS change 
than ati-vga.

Regards,
BALATON Zoltan