Re: BPF: failed module verification on linux-next

[Andrii Nakryiko]

On Mon, May 24, 2021 at 3:58 PM Andrii Nakryiko
<andrii.nakryiko@xxxxxxxxx> wrote:
>
> On Thu, May 20, 2021 at 10:31 PM Andrii Nakryiko
> <andrii.nakryiko@xxxxxxxxx> wrote:
> >
> > On Wed, May 19, 2021 at 7:19 AM Michal Suchánek <msuchanek@xxxxxxx> wrote:
> > >
> > > Hello,
> > >
> > > linux-next fails to boot for me:
> > >
> > > [    0.000000] Linux version 5.13.0-rc2-next-20210519-1.g3455ff8-vanilla (geeko@buildhost) (gcc (SUSE Linux) 10.3.0, GNU ld (GNU Binutils;
> > > openSUSE Tumbleweed) 2.36.1.20210326-3) #1 SMP Wed May 19 10:05:10 UTC 2021 (3455ff8)
> > > [    0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.13.0-rc2-next-20210519-1.g3455ff8-vanilla root=UUID=ec42c33e-a2c2-4c61-afcc-93e9527
> > > 8f687 plymouth.enable=0 resume=/dev/disk/by-uuid/f1fe4560-a801-4faf-a638-834c407027c7 mitigations=auto earlyprintk initcall_debug nomodeset
> > >  earlycon ignore_loglevel console=ttyS0,115200
> > > ...
> > > [   26.093364] calling  tracing_set_default_clock+0x0/0x62 @ 1
> > > [   26.098937] initcall tracing_set_default_clock+0x0/0x62 returned 0 after 0 usecs
> > > [   26.106330] calling  acpi_gpio_handle_deferred_request_irqs+0x0/0x7c @ 1
> > > [   26.113033] initcall acpi_gpio_handle_deferred_request_irqs+0x0/0x7c returned 0 after 3 usecs
> > > [   26.121559] calling  clk_disable_unused+0x0/0x102 @ 1
> > > [   26.126620] initcall clk_disable_unused+0x0/0x102 returned 0 after 0 usecs
> > > [   26.133491] calling  regulator_init_complete+0x0/0x25 @ 1
> > > [   26.138890] initcall regulator_init_complete+0x0/0x25 returned 0 after 0 usecs
> > > [   26.147816] Freeing unused decrypted memory: 2036K
> > > [   26.153682] Freeing unused kernel image (initmem) memory: 2308K
> > > [   26.165776] Write protecting the kernel read-only data: 26624k
> > > [   26.173067] Freeing unused kernel image (text/rodata gap) memory: 2036K
> > > [   26.180416] Freeing unused kernel image (rodata/data gap) memory: 1184K
> > > [   26.187031] Run /init as init process
> > > [   26.190693]   with arguments:
> > > [   26.193661]     /init
> > > [   26.195933]   with environment:
> > > [   26.199079]     HOME=/
> > > [   26.201444]     TERM=linux
> > > [   26.204152]     BOOT_IMAGE=/boot/vmlinuz-5.13.0-rc2-next-20210519-1.g3455ff8-vanilla
> > > [   26.254154] BPF:      type_id=35503 offset=178440 size=4
> > > [   26.259125] BPF:
> > > [   26.261054] BPF:Invalid offset
> > > [   26.264119] BPF:
> >
> > It took me a while to reliably bisect this, but it clearly points to
> > this commit:
> >
> > e481fac7d80b ("mm/page_alloc: convert per-cpu list protection to local_lock")
> >
> > One commit before it, 676535512684 ("mm/page_alloc: split per cpu page
> > lists and zone stats -fix"), works just fine.
> >
> > I'll have to spend more time debugging what exactly is happening, but
> > the immediate problem is two different definitions of numa_node
> > per-cpu variable. They both are at the same offset within
> > .data..percpu ELF section, they both have the same name, but one of
> > them is marked as static and another as global. And one is int
> > variable, while another is struct pagesets. I'll look some more
> > tomorrow, but adding Jiri and Arnaldo for visibility.
> >
> > [110907] DATASEC '.data..percpu' size=178904 vlen=303
> > ...
> >         type_id=27753 offset=163976 size=4 (VAR 'numa_node')
> >         type_id=27754 offset=163976 size=4 (VAR 'numa_node')
> >
> > [27753] VAR 'numa_node' type_id=27556, linkage=static
> > [27754] VAR 'numa_node' type_id=20, linkage=global
> >
> > [20] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED
> >
> > [27556] STRUCT 'pagesets' size=0 vlen=1
> >         'lock' type_id=507 bits_offset=0
> >
> > [506] STRUCT '(anon)' size=0 vlen=0
> > [507] TYPEDEF 'local_lock_t' type_id=506
> >
> > So also something weird about those zero-sized struct pagesets and
> > local_lock_t inside it.
>
> Ok, so nothing weird about them. local_lock_t is designed to be
> zero-sized unless CONFIG_DEBUG_LOCK_ALLOC is defined.
>
> But such zero-sized per-CPU variables are confusing pahole during BTF
> generation, as now two different variables "occupy" the same address.

FWIW, here's the pahole fix (it tried to filter zero-sized per-CPU
vars, but not quite completely).

  [0] https://lore.kernel.org/bpf/20210524234222.278676-1-andrii@xxxxxxxxxx/T/#u

>
> Given this seems to be the first zero-sized per-CPU variable, I wonder
> if it would be ok to make sure it's never zero-sized, while pahole
> gets fixed and it's latest version gets widely packaged and
> distributed.
>
> Mel, what do you think about something like below? Or maybe you can
> advise some better solution?
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 41b87d6f840c..6a1d7511cae9 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -124,6 +124,13 @@ static DEFINE_MUTEX(pcp_batch_high_lock);
>
>  struct pagesets {
>      local_lock_t lock;
> +#if defined(CONFIG_DEBUG_INFO_BTF) && !defined(CONFIG_DEBUG_LOCK_ALLOC)
> +    /* pahole 1.21 and earlier gets confused by zero-sized per-CPU
> +     * variables and produces invalid BTF. So to accommodate earlier
> +     * versions of pahole, ensure that sizeof(struct pagesets) is never 0.
> +     */
> +    char __filler;
> +#endif
>  };
>  static DEFINE_PER_CPU(struct pagesets, pagesets) = {
>      .lock = INIT_LOCAL_LOCK(lock),
>
> >
> > > [   26.264119]
> > > [   26.267437] failed to validate module [efivarfs] BTF: -22
> > > [   26.316724] systemd[1]: systemd 246.13+suse.105.g14581e0120 running in system mode. (+PAM +AUDIT +SELINUX -IMA +APPARMOR -SMACK +SYSVINI
> > > T +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +ZSTD +SECCOMP +BLKID +ELFUTILS +KMOD +IDN2 -IDN +PCRE2 default-hierarchy=unified)
> > > [   26.357990] systemd[1]: Detected architecture x86-64.
> > > [   26.363068] systemd[1]: Running in initial RAM disk.
> > >
> >
> > [...]