Opened 19 months ago

Last modified 19 months ago

#18022 new bug

SD card causes kernel panic

Reported by: Parodper Owned by: pulkomandy
Priority: normal Milestone: Unscheduled
Component: Drivers/Disk/MMC Version: R1/Development
Keywords: Cc:
Blocked By: Blocking:
Platform: All

Description

Immediately after putting in the SD card in the reader (see the end of the listdev.txt for more info) I get the following kernel panic:

PANIC: Command execution unwise, data inhibit

Welcome to Kernel Debugging Land...
Thread 442 "mmc storage scheduler 2" running on CPU 1
stack trace for thread 442 "mmc storage scheduler 2"
    kernel stack: 0xffffffff81b72000 to 0xffffffff81b77000
frame                          caller             <image>:function + offset
 0 0xffffffff81b76b48 (+  24)  ffffffff8014550c   <kernel_x86_64> arch_debug_call_with_fault_handler + 0x16
 1 0xffffffff81b76b60 (+  80)  ffffffff800aec18   <kernel_x86_64> debug_call_with_fault_handler + 0x78
 2 0xffffffff81b76bb0 (+  96)  ffffffff800b0233   <kernel_x86_64> kernel_debugger_loop (char const*, char const*, __va_list_tag*, int) + Oxf3
 3 0xffffffff81b76c10 (+  80)  ffffffff800b05ce   <kernel_x86_64> kernel_debugger internal (char const*, char const*, __va_list_tag*, int) + Ox6e
 4 0xffffffff81b76c60 (+ 240)  ffffffff800b0927   <kernel_x86_64> panic + 0xb7
 5 0xffffffff81b76d50 (+  80)  ffffffff819c1fc3   <sdhci_pci> SdhciBus::ExecuteCommand[clone .localalias] (unsigned char, unsigned int, unsigned int*) + 0x363
 6 0xffffffff81b76da0 (+ 112)  ffffffff819c2342   <sdhci_pci> SdhciBus::DoIO[clone .localalias] (unsigned char, IOOperation*, bool) + 0x132
 7 0xffffffff81b76e10 (+  64)  ffffffff816881df   <mmc> mmc_bus_do_io(device_node*, void*, unsigned short, unsigned char, IOOperation*, bool) + 0x3f
 8 0xffffffff81b76e50 (+  48)  ffffffff81b3a521   </boot/system/add-ons/kernel/drivers/disk/mmc/mmc_disk> mmc_disk_execute_iorequest(void*, IOOperation*) + 0x41
 9 0xffffffff81b76e80 (+ 304)  ffffffff800d67b5   <kernel_x86_64> IOSchedulerSimple::__Scheduler() + 0x415
10 0xffffffff81b76fb0 (+  32)  ffffffff8008c848   <kernel_x86_64>commor_thread_entry(void*) + 0x38
11 0xffffffff81b76fd0 (+2118684720)  ffffffff81a31fe0   7523:mmc storage scheduler 2_441_kst@0xffffffff81a2d000 + 0x4fe0

I can reproduce this on QEMU, with PCI passthrough. Using hrev56555 on x86_64.

Attachments (1)

listdev.txt (5.3 KB ) - added by Parodper 19 months ago.
Output of listdev on Haiku and lspci on Linux

Download all attachments as: .zip

Change History (5)

by Parodper, 19 months ago

Attachment: listdev.txt added

Output of listdev on Haiku and lspci on Linux

comment:1 by diver, 19 months ago

Keywords: mmc removed
Version: R1/beta3R1/Development

comment:2 by Parodper, 19 months ago

PD: For the life of me I couldn't find out how to use generate the QR code.

comment:3 by waddlesplash, 19 months ago

QR codes are only relevant as a way to get copy/pastable data. I see you've retyped the KDL (as I can see typos in it), if you can reproduce the issue in QEMU then you can just use -serial stdio and then copy the panic trace from there.

comment:4 by Parodper, 19 months ago

Thanks for the tip, but I first tried the QR thing on real hardware.

Note: See TracTickets for help on using tickets.