1. Patchew
  2. linux
  3. View series

[PATCH v2] bpf: Fix bpf_get_smp_processor_id() on !CONFIG_SMP

Andrea Righi posted 1 patch 1 year ago
There is a newer version of this series
kernel/bpf/verifier.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
[PATCH v2] bpf: Fix bpf_get_smp_processor_id() on !CONFIG_SMP
Posted by Andrea Righi 1 year ago 
On x86-64 calling bpf_get_smp_processor_id() in a kernel with CONFIG_SMP
disabled can trigger the following bug, as pcpu_hot is unavailable:

 [    8.471774] BUG: unable to handle page fault for address: 00000000936a290c
 [    8.471849] #PF: supervisor read access in kernel mode
 [    8.471881] #PF: error_code(0x0000) - not-present page

Fix by inlining a return 0 in the !CONFIG_SMP case.

Fixes: 1ae6921009e5 ("bpf: inline bpf_get_smp_processor_id() helper")
Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
---
 kernel/bpf/verifier.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

ChangeLog v1 -> v2:
  - inline a "return 0" instead of not inlining bpf_get_smp_processor_id() at
    all in the !CONFIG_SMP case, as suggested by Daniel

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index f7f892a52a37..761c70899754 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -21281,11 +21281,15 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
 			 * changed in some incompatible and hard to support
 			 * way, it's fine to back out this inlining logic
 			 */
+#ifdef CONFIG_SMP
 			insn_buf[0] = BPF_MOV32_IMM(BPF_REG_0, (u32)(unsigned long)&pcpu_hot.cpu_number);
 			insn_buf[1] = BPF_MOV64_PERCPU_REG(BPF_REG_0, BPF_REG_0);
 			insn_buf[2] = BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_0, 0);
 			cnt = 3;
-
+#else
+			BPF_ALU32_REG(BPF_XOR, BPF_REG_0, BPF_REG_0),
+			cnt = 1;
+#endif
 			new_prog = bpf_patch_insn_data(env, i + delta, insn_buf, cnt);
 			if (!new_prog)
 				return -ENOMEM;
-- 
2.47.1
Re: [PATCH v2] bpf: Fix bpf_get_smp_processor_id() on !CONFIG_SMP
Posted by Andrii Nakryiko 1 year ago 
On Mon, Dec 16, 2024 at 1:00 PM Andrea Righi <arighi@nvidia.com> wrote:
>
> On x86-64 calling bpf_get_smp_processor_id() in a kernel with CONFIG_SMP
> disabled can trigger the following bug, as pcpu_hot is unavailable:
>
>  [    8.471774] BUG: unable to handle page fault for address: 00000000936a290c
>  [    8.471849] #PF: supervisor read access in kernel mode
>  [    8.471881] #PF: error_code(0x0000) - not-present page
>
> Fix by inlining a return 0 in the !CONFIG_SMP case.
>
> Fixes: 1ae6921009e5 ("bpf: inline bpf_get_smp_processor_id() helper")
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Signed-off-by: Andrea Righi <arighi@nvidia.com>
> ---
>  kernel/bpf/verifier.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
>
> ChangeLog v1 -> v2:
>   - inline a "return 0" instead of not inlining bpf_get_smp_processor_id() at
>     all in the !CONFIG_SMP case, as suggested by Daniel
>
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index f7f892a52a37..761c70899754 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -21281,11 +21281,15 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
>                          * changed in some incompatible and hard to support
>                          * way, it's fine to back out this inlining logic
>                          */
> +#ifdef CONFIG_SMP
>                         insn_buf[0] = BPF_MOV32_IMM(BPF_REG_0, (u32)(unsigned long)&pcpu_hot.cpu_number);
>                         insn_buf[1] = BPF_MOV64_PERCPU_REG(BPF_REG_0, BPF_REG_0);
>                         insn_buf[2] = BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_0, 0);
>                         cnt = 3;
> -
> +#else
> +                       BPF_ALU32_REG(BPF_XOR, BPF_REG_0, BPF_REG_0),

um... shouldn't this be `insns_buf[0] = ` assignment? And that comma
instead of semicolon at the end?

pw-bot: cr

> +                       cnt = 1;
> +#endif
>                         new_prog = bpf_patch_insn_data(env, i + delta, insn_buf, cnt);
>                         if (!new_prog)
>                                 return -ENOMEM;
> --
> 2.47.1
>
Re: [PATCH v2] bpf: Fix bpf_get_smp_processor_id() on !CONFIG_SMP
Posted by Andrea Righi 1 year ago 
On Tue, Dec 17, 2024 at 10:38:04AM -0800, Andrii Nakryiko wrote:
...
> > diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> > index f7f892a52a37..761c70899754 100644
> > --- a/kernel/bpf/verifier.c
> > +++ b/kernel/bpf/verifier.c
> > @@ -21281,11 +21281,15 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
> >                          * changed in some incompatible and hard to support
> >                          * way, it's fine to back out this inlining logic
> >                          */
> > +#ifdef CONFIG_SMP
> >                         insn_buf[0] = BPF_MOV32_IMM(BPF_REG_0, (u32)(unsigned long)&pcpu_hot.cpu_number);
> >                         insn_buf[1] = BPF_MOV64_PERCPU_REG(BPF_REG_0, BPF_REG_0);
> >                         insn_buf[2] = BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_0, 0);
> >                         cnt = 3;
> > -
> > +#else
> > +                       BPF_ALU32_REG(BPF_XOR, BPF_REG_0, BPF_REG_0),
> 
> um... shouldn't this be `insns_buf[0] = ` assignment? And that comma
> instead of semicolon at the end?

Yeah.. my bad, I tested it with the wrong .config that has CONFIG_SMP
enabled.

I'll send a v3 with the proper code, sorry for the noise.

-Andrea

> 
> pw-bot: cr
> 
> > +                       cnt = 1;
> > +#endif
> >                         new_prog = bpf_patch_insn_data(env, i + delta, insn_buf, cnt);
> >                         if (!new_prog)
> >                                 return -ENOMEM;
> > --
> > 2.47.1
> >

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.