Re: [PATCH bpf-next v3 1/2] x86: Perform BPF exception fixup in do_user_addr_fault

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 11/4/24 09:50, Kumar Kartikeya Dwivedi wrote:
> While reading invalid memory is a rare case that should not happen,
> it may be possible if some kernel field contains a stale address
> etc.

Reading random (unaccepted) memory in a TDX guest is fatal, even from
kernel addresses.  We had a lot of fun even getting
load_unaligned_zeropad() to work.

So, honestly, if you've letting buggy BFP programs get loaded and read
random memory (kernel or user), you've got bigger problems than a
verbose kernel panic when the buggy program happens to touch userspace.

I'd rather not hack code into the page fault handler to add to the
illusion that this is a good idea or safe in any way.




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux