Re: [PATCH 1/2] libtraceevent: Add eof checks.

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

 



On Wed, Jun 23, 2021 at 6:06 PM Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
>
> On Thu, 17 Jun 2021 15:58:28 -0400
> Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
>
> > On Thu, 17 Jun 2021 19:43:25 +0000
> > Claire Jensen <cjense@xxxxxxxxxx> wrote:
> >
> > Hi Claire,
> >
> > Thanks for sending the patches, I'll try to get some time to look at them
> > (note, that I have a lot of other duties that I need to finish before I can
> > get to this).
> >
> > > Added checking for __read_char and peek_char to make sure value is not at end
> > > of file.
> > >
> > > This issue was found while fuzz testing. One of the test cases created an infinite loop because __read_token had reached end of file. Checking was added to all cases where this may occur.
> >
> > You don't need to fix this now, but for future reference, we follow the
> > Linux guidelines on submitting patches which includes having line breaks at
> > 74 (although I use 76) bytes, for the change log.
> >
> > This makes the change logs easier to read.
> >
>
> I made the mistake of adding this patch and pushing it to a new release
> without running my test suite against it. It ended up breaking the parsing.
>
> When running with --debug -N, I get:

Hi Steve,

sorry for the breakage, could you give a full reproduction command?

Thanks!
Ian

>   [ftrace:branch] unexpected type 1
>   [sched:sched_switch] unknown op ''
>   [irq:irq_handler_exit] unexpected type 1
>   [timer:timer_start] unknown op ''
>   [kvm:vcpu_match_mmio] unexpected type 1
>   [kvm:kvm_wait_lapic_expire] unknown op ''
>   [kvm:kvm_vcpu_wakeup] unexpected type 1
>   [kvm:kvm_userspace_exit] unknown op ''
>   [kvm:kvm_pv_tlb_flush] unexpected type 1
>   [kvm:kvm_ple_window_update] unknown op ''
>   [kvm:kvm_pio] unknown op ''
>   [kvm:kvm_pic_set_irq] unknown op ''
>   [kvm:kvm_nested_vmrun] unexpected type 1
>   [kvm:kvm_nested_vmexit_inject] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_nested_vmexit] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_nested_vmenter_failed] bad op token
>   [kvm:kvm_msr] unexpected type 1
>   [kvm:kvm_msi_set_irq] unknown op ''
>   unknown op ''
>   [kvm:kvm_ioapic_set_irq] unknown op ''
>   Error: expected type 5 but read 0
>   unknown op ''
>   [kvm:kvm_ioapic_delayed_eoi_inj] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_exit] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_emulate_insn] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_cpuid] unexpected type 1
>   unknown op ''
>   [kvm:kvm_apic_ipi] unknown op ''
>   Error: expected type 5 but read 0
>   unknown op ''
>   [kvm:kvm_apic_accept_irq] unknown op ''
>   Error: expected type 5 but read 0
>   [kvm:kvm_age_page] unexpected type 1
>
> with the patch, and no errors without it.
>
> I have to revert this patch and push a new version out.
>
> -- Steve



[Index of Archives]     [Linux USB Development]     [Linux USB Development]     [Linux Audio Users]     [Yosemite Hiking]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux