Opened 11 months ago

Last modified 11 months ago

#18468 new bug

Regression in VFS

Reported by: kallisti5 Owned by: nobody
Priority: critical Milestone: Unscheduled
Component: Drivers/Disk/USB Version: R1/Development
Keywords: Cc:
Blocked By: Blocking:
Platform: riscv64

Description

Attempting to get data from a usb-disk... 

cp /USBDISK/binutils... ~/
cp: error reading /USBDISK/binutils...: Bad address

usb_disk: failed to lock memory: Bad address
file_cache: read pages failed: Bad address
(repeat)

seen on hrev57100. Not seen on hrev57093

Change History (5)

comment:1 by waddlesplash, 11 months ago

Component: - General → Drivers/Disk/USB

This is a strange one.

comment:2 by kallisti5, 11 months ago

full disclosure, this was on riscv64 on my unmatched.. however this is a new issue on that host with the upgrade listed (same USB flash disk, same USB port). Tested FAT and NTFS filesystems, saw the same issue.

I get a feeling that x86 will see it as well.

comment:3 by pulkomandy, 11 months ago

Platform: All → riscv64

Marking as platform: riscv64 until proven otherwise :)

comment:4 by korli, 11 months ago

Version: R1/beta4 → R1/Development

comment:5 by kallisti5, 11 months ago

eh. ok. I upgraded my x86 desktop and attempted the same transaction on the same flash drive. Couldn't reproduce the issue. I gotta assume at this point it wasn't related to the VFS changes, but related to the local versions of gcc_syslibs I have installed. We're running a weird mix of gcc_syslibs trying to get a fresh gcc hpkg

Leaving open for now, but i'll close once confirmed.

Note: See TracTickets for help on using tickets.