Re: [PATCH bpf-next v1 0/4] libbpf: Support kernel-side kfunc flavors
From: Ihor Solodrai
Date: Wed Nov 18 2026 - 16:37:49 EST
On 2026-11-18 2:12 p.m., Eduard Zingerman wrote:
On Tue, 2026-11-17 at 12:21 -0700, Ihor Solodrai wrote:
An incompatible change to a kfunc's prototype, such as an added
parameter, breaks BPF programs built against the old one. libbpf
ignores the ___<flavor> suffix of a kfunc declared in a BPF program,
but makes an exact name look up on the kernel side, so the kernel
can't keep the old prototype available under the kfunc's name.
sched_ext ran into this. cded46d97159 ("sched_ext: Make
scx_bpf_dsq_insert*() return bool") kept the old prototype as
scx_bpf_dsq_insert___compat, expecting libbpf to bind old binaries to
it, and broke them. 2dbbdeda77a6 ("sched_ext: Fix scx_bpf_dsq_insert()
backward binary compatibility") reversed the naming. A ___v2 kfunc
declared as vmlinux.h spells it doesn't bind: 56bbc91219c0
("sched_ext: Fix scx_bpf_dsq_move_to_local___v2 compat detection")
fixed a compat macro that silently dropped enq_flags because of this.
Please prefer examples to prose in cover letters.
The sched_ext story above narrates the commit history to illustrate
the motivation for the feature. Would you like to see code samples
from there, or are you suggesting to drop the paragraph?
The documentation patch includes code examples of the new binding
rules, I didn't want to duplicate that in the cover.
Also, please de-Claude the included prose.
Please be more specific. Which "included prose"?
With this series, if the kernel's bpf_foo kfunc is missing or its
prototype is incompatible with the declaration, libbpf binds the
bpf_foo___<flavor> that has a compatible prototype. So a changed kfunc
can keep its name, and the old prototype stays available as foo___v1.
resolve_btfids fails the build if two of bpf_foo and its flavors have
compatible prototypes, unless both have foo's exact prototype
(aliases), so the fallback finds at most one match. This ensures at
Why is an exception for kfuncs with identical prototypes necessary?
For example, when the void scx_bpf_dsq_insert() is deprecated, the
kernel side can do this:
Today: void scx_bpf_dsq_insert(...)
bool scx_bpf_dsq_insert___v2(...)
Later: bool scx_bpf_dsq_insert(...)
bool scx_bpf_dsq_insert___v2(...) /* calls scx_bpf_dsq_insert() */
Existing schedulers declare scx_bpf_dsq_insert___v2___compat.
libbpf strips the last suffix, so the bpf progs still call
scx_bpf_dsq_insert___v2(). The wrapper keeps them working during the
transition, including with older libbpf.
The exception allows to do this.
Otherwise we break the ___v2___compat trick.
kernel / module build time that no ambigous kfunc flavors exist.
Documentation/bpf/kfuncs.rst describes the rules.
Compatibility is determined by bpf_core_types_are_compat(), the check
libbpf already applies to kfuncs. Of the 70 same-name kfunc prototype
changes between v6.6 and bpf-next, it finds 8 incompatible (arity,
return kind, callback prototype). The other 62 (const, typedefs,
parameter names, integer widths, pointee struct names, annotation
suffixes) were made in place without breaking old callers.
Do you mean that without this series programs referring old prototypes
are broken?
Yes, if the kernel changed the prototype in place and the BPF program
still expects the old one. For example, 8fe4dc4f6456 ("bpf: change
prototype of bpf_session_{cookie,is_return}") added a ctx argument to
bpf_session_is_return() in v7.0. libbpf can't resolve an object's old
no-argument declaration on that kernel.
This series alone doesn't fix this, but with it the kernel *can*
provide bpf_session_is_return___v1(void) for the old callers.
...