| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <f6fd8209-4808-45e1-b5b5-95ea2dcc0ba4@bytedance.com>
Date: Wed, 11 Oct 2023 12:44:08 +0800
From: Chuyi Zhou <zhouchuyi@...edance.com>
To: bpf@...r.kernel.org
Cc: ast@...nel.org, daniel@...earbox.net, andrii@...nel.org,
martin.lau@...nel.org, tj@...nel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH bpf-next v4 4/8] bpf: Introduce css open-coded iterator
kfuncs
在 2023/10/7 20:45, Chuyi Zhou 写道:
> This Patch adds kfuncs bpf_iter_css_{new,next,destroy} which allow
> creation and manipulation of struct bpf_iter_css in open-coded iterator
> style. These kfuncs actually wrapps css_next_descendant_{pre, post}.
> css_iter can be used to:
>
> 1) iterating a sepcific cgroup tree with pre/post/up order
>
> 2) iterating cgroup_subsystem in BPF Prog, like
> for_each_mem_cgroup_tree/cpuset_for_each_descendant_pre in kernel.
>
> The API design is consistent with cgroup_iter. bpf_iter_css_new accepts
> parameters defining iteration order and starting css. Here we also reuse
> BPF_CGROUP_ITER_DESCENDANTS_PRE, BPF_CGROUP_ITER_DESCENDANTS_POST,
> BPF_CGROUP_ITER_ANCESTORS_UP enums.
>
> Signed-off-by: Chuyi Zhou <zhouchuyi@...edance.com>
> Acked-by: Tejun Heo <tj@...nel.org>
> ---
> kernel/bpf/cgroup_iter.c | 59 +++++++++++++++++++
> kernel/bpf/helpers.c | 3 +
> .../testing/selftests/bpf/bpf_experimental.h | 6 ++
> 3 files changed, 68 insertions(+)
>
> diff --git a/kernel/bpf/cgroup_iter.c b/kernel/bpf/cgroup_iter.c
> index 810378f04fbc..9c6ad892ae82 100644
> --- a/kernel/bpf/cgroup_iter.c
> +++ b/kernel/bpf/cgroup_iter.c
> @@ -294,3 +294,62 @@ static int __init bpf_cgroup_iter_init(void)
> }
>
> late_initcall(bpf_cgroup_iter_init);
> +
> +struct bpf_iter_css {
> + __u64 __opaque[3];
> +} __attribute__((aligned(8)));
> +
> +struct bpf_iter_css_kern {
> + struct cgroup_subsys_state *start;
> + struct cgroup_subsys_state *pos;
> + unsigned int flags;
> +} __attribute__((aligned(8)));
> +
> +__bpf_kfunc int bpf_iter_css_new(struct bpf_iter_css *it,
> + struct cgroup_subsys_state *start, unsigned int flags)
> +{
> + struct bpf_iter_css_kern *kit = (void *)it;
> +
> + BUILD_BUG_ON(sizeof(struct bpf_iter_css_kern) != sizeof(struct bpf_iter_css));
> + BUILD_BUG_ON(__alignof__(struct bpf_iter_css_kern) != __alignof__(struct bpf_iter_css));
> +
This would cause the fail of netdev/build_32bit CI
(https://netdev.bots.linux.dev/static/nipa/790929/13412333/build_32bit/stderr):
tools/testing/selftests/kvm/settings: warning: ignored by one of the
.gitignore files
../kernel/bpf/cgroup_iter.c:308:17: warning: no previous prototype for
‘bpf_iter_css_new’ [-Wmissing-prototypes]
308 | __bpf_kfunc int bpf_iter_css_new(struct bpf_iter_css *it,
| ^~~~~~~~~~~~~~~~
../kernel/bpf/cgroup_iter.c:332:41: warning: no previous prototype for
‘bpf_iter_css_next’ [-Wmissing-prototypes]
332 | __bpf_kfunc struct cgroup_subsys_state
*bpf_iter_css_next(struct bpf_iter_css *it)
| ^~~~~~~~~~~~~~~~~
../kernel/bpf/cgroup_iter.c:353:18: warning: no previous prototype for
‘bpf_iter_css_destroy’ [-Wmissing-prototypes]
353 | __bpf_kfunc void bpf_iter_css_destroy(struct bpf_iter_css *it)
| ^~~~~~~~~~~~~~~~~~~~
In file included from <command-line>:
../kernel/bpf/cgroup_iter.c: In function ‘bpf_iter_css_new’:
./../include/linux/compiler_types.h:425:45: error: call to
‘__compiletime_assert_322’ declared with attribute error: BUILD_BUG_ON
failed: sizeof(struct bpf_iter_css_kern) != sizeof(struct bpf_iter_css)
425 | _compiletime_assert(condition, msg,
__compiletime_assert_, __COUNTER__)
| ^
./../include/linux/compiler_types.h:406:25: note: in definition of macro
‘__compiletime_assert’
406 | prefix ## suffix();
\
| ^~~~~~
./../include/linux/compiler_types.h:425:9: note: in expansion of macro
‘_compiletime_assert’
425 | _compiletime_assert(condition, msg,
__compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
../include/linux/build_bug.h:39:37: note: in expansion of macro
‘compiletime_assert’
39 | #define BUILD_BUG_ON_MSG(cond, msg) compiletime_assert(!(cond),
msg)
| ^~~~~~~~~~~~~~~~~~
../include/linux/build_bug.h:50:9: note: in expansion of macro
‘BUILD_BUG_ON_MSG’
50 | BUILD_BUG_ON_MSG(condition, "BUILD_BUG_ON failed: "
#condition)
| ^~~~~~~~~~~~~~~~
../kernel/bpf/cgroup_iter.c:313:9: note: in expansion of macro
‘BUILD_BUG_ON’
313 | BUILD_BUG_ON(sizeof(struct bpf_iter_css_kern) !=
sizeof(struct bpf_iter_css));
The reason seems on 32-bit machine, sizeof(struct bpf_iter_css) is 24
and sizeof(struct bpf_iter_css_kern) is 16.
I was wondering whether the BUILD_BUG_ON check is necessary. Looking at
the struct bpf_list_node and struct bpf_list_node_kern wich are very
similay to bpf_iter_css, I didn't see the BUILD_BUG_ON check when
convert from (struct bpf_list_node *) to (struct bpf_list_node_kern *)
/* Non-opaque version of bpf_list_node in uapi/linux/bpf.h */
struct bpf_list_node_kern {
struct list_head list_head;
void *owner;
} __attribute__((aligned(8)));
struct bpf_list_node {
__u64 :64;
__u64 :64;
__u64 :64;
} __attribute__((aligned(8)));
__bpf_kfunc int bpf_list_push_back_impl(struct bpf_list_head *head,
struct bpf_list_node *node,
void *meta__ign, u64 off)
{
struct bpf_list_node_kern *n = (void *)node;
}
Powered by blists - more mailing lists