[PATCH v5 0/3] fanotify: Allow user space to pass back additional audit info
by Richard Guy Briggs
The Fanotify API can be used for access control by requesting permission
event notification. The user space tooling that uses it may have a
complicated policy that inherently contains additional context for the
decision. If this information were available in the audit trail, policy
writers can close the loop on debugging policy. Also, if this additional
information were available, it would enable the creation of tools that
can suggest changes to the policy similar to how audit2allow can help
refine labeled security.
This patchset defines a new flag (FAN_INFO) and new extensions that
define additional information which are appended after the response
structure returned from user space on a permission event. The appended
information is organized with headers containing a type and size that
can be delegated to interested subsystems. One new information type is
defined to audit the triggering rule number.
A newer kernel will work with an older userspace and an older kernel
will behave as expected and reject a newer userspace, leaving it up to
the newer userspace to test appropriately and adapt as necessary.
The audit function was updated to log the additional information in the
AUDIT_FANOTIFY record. The following are examples of the new record
format:
type=FANOTIFY msg=audit(1600385147.372:590): resp=2 fan_type=1 fan_info=3137 subj_trust=3 obj_trust=5
type=FANOTIFY msg=audit(1659730979.839:284): resp=1 fan_type=0 fan_info=3F subj_trust=? obj_trust=?
changelog:
v1:
- first version by Steve Grubb <sgrubb(a)redhat.com>
Link: https://lore.kernel.org/r/2042449.irdbgypaU6@x2
v2:
- enhancements suggested by Jan Kara <jack(a)suse.cz>
- 1/3 change %d to %u in pr_debug
- 2/3 change response from __u32 to __u16
- mod struct fanotify_response and fanotify_perm_event add extra_info_type, extra_info_buf
- extra_info_buf size max FANOTIFY_MAX_RESPONSE_EXTRA_LEN, add struct fanotify_response_audit_rule
- extend debug statements
- remove unneeded macros
- [internal] change interface to finish_permission_event() and process_access_response()
- 3/3 update format of extra information
- [internal] change interface to audit_fanotify()
- change ctx_type= to fan_type=
Link: https://lore.kernel.org/r/cover.1651174324.git.rgb@redhat.com
v3:
- 1/3 switch {,__}audit_fanotify() from uint to u32
- 2/3 re-add fanotify_get_response switch case FAN_DENY: to avoid unnecessary churn
- add FAN_EXTRA flag to indicate more info and break with old kernel
- change response from u16 to u32 to avoid endian issues
- change extra_info_buf to union
- move low-cost fd check earlier
- change FAN_RESPONSE_INFO_AUDIT_NONE to FAN_RESPONSE_INFO_NONE
- switch to u32 for internal and __u32 for uapi
Link: https://lore.kernel.org/all/cover.1652730821.git.rgb@redhat.com
v4:
- scrap FAN_INVALID_RESPONSE_MASK in favour of original to catch invalid response == 0
- introduce FANOTIFY_RESPONSE_* macros
- uapi: remove union
- keep original struct fanotify_response, add fan_info infra starting with audit reason
- uapi add struct fanotify_response_info_header{type/pad/len} and struct fanotify_response_info_audit_rule{hdr/rule}
- rename fan_ctx= to fan_info=, FAN_EXTRA to FAN_INFO
- change event struct from type/buf to len/buf
- enable multiple info extensions in one message
- hex encode fan_info in __audit_fanotify()
- record type FANOTIFY extended to "type=FANOTIFY msg=audit(1659730979.839:284): resp=1 fan_type=0 fan_info=3F"
Link: https://lore.kernel.org/all/cover.1659996830.git.rgb@redhat.com
v5:
- fixed warnings in p2/4 and p3/4 found by <lkp(a)intel.com>
- restore original behaviour for !FAN_INFO case and fanotify_get_response()
- rename member audit_rule to rule_number
- eliminate memory leak of info_buf on failure (no longer dynamic)
- rename buf:info, count:info_len, c:remain, ib:infop
- fix pr_debug
- return -ENOENT on FAN_INFO and fd==FAN_NOFD to signal new kernel
- fanotify_write() remove redundant size check
- add u32 subj_trust obj_trust fields with unknown value "2"
- split out to helper process_access_response_info()
- restore finish_permission_event() response_struct to u32
- assume and enforce one rule to audit, pass struct directly to __audit_fanotify()
- change fanotify_perm_event struct to union hdr/audir_rule
- add vspace to fanotify_write() and process_access_response_info()
- squash 3/4 with 4/4
- fix v3 and v4 links
Link: https://lore.kernel.org/all/cover.1670606054.git.rgb@redhat.com
Richard Guy Briggs (3):
fanotify: Ensure consistent variable type for response
fanotify: define struct members to hold response decision context
fanotify,audit: Allow audit to use the full permission event response
fs/notify/fanotify/fanotify.c | 8 ++-
fs/notify/fanotify/fanotify.h | 6 +-
fs/notify/fanotify/fanotify_user.c | 88 ++++++++++++++++++++++--------
include/linux/audit.h | 9 +--
include/linux/fanotify.h | 5 ++
include/uapi/linux/fanotify.h | 30 +++++++++-
kernel/auditsc.c | 25 ++++++++-
7 files changed, 138 insertions(+), 33 deletions(-)
--
2.27.0
1 year, 11 months
Re: [RFC PATCH 00/25] Upstream kvx Linux port
by Jeff Xie
On Mon, Jan 9, 2023 at 11:53 PM Jeff Xie <xiehuan09(a)gmail.com> wrote:
>
> On Mon, Jan 9, 2023 at 11:30 PM Yann Sionneau <ysionneau(a)kalray.eu> wrote:
> >
> > Hi Jeff,
> >
> > On 1/9/23 16:11, Jeff Xie wrote:
> > > On Mon, Jan 9, 2023 at 9:21 PM Yann Sionneau <ysionneau(a)kalray.eu> wrote:
> > >> Hi Jeff,
> > >>
> > >> On 1/7/23 07:25, Jeff Xie wrote:
> > >>> Hi,
> > >>>
> > >>> On Wed, Jan 4, 2023 at 1:01 AM Yann Sionneau <ysionneau(a)kalray.eu> wrote:
> > >>>> [snip]
> > >>>>
> > >>>> A kvx toolchain can be built using:
> > >>>> # install dependencies: texinfo bison flex libgmp-dev libmpc-dev libmpfr-dev
> > >>>> $ git clone https://github.com/kalray/build-scripts
> > >>>> $ cd build-scripts
> > >>>> $ source last.refs
> > >>>> $ ./build-kvx-xgcc.sh output
> > >>> I would like to build the kvx-xgcc to compile and test the linux
> > >>> kernel, but it reported a compile error.
> > >>> I wonder what version of gcc you are using.
> > >>>
> > >>> My build environment:
> > >>> VERSION="20.04.2 LTS (Focal Fossa)"
> > >>> gcc version 9.3.0 (Ubuntu 9.3.0-17ubuntu1~20.04)
> > >>>
> > >>>
> > >>> Compile error:
> > >>> $ ./build-kvx-xgcc.sh output
> > >>>
> > >>> ../../binutils/libiberty/fibheap.c: In function ‘fibheap_replace_key_data’:
> > >>> ../../binutils/libiberty/fibheap.c:38:24: error: ‘LONG_MIN’ undeclared
> > >>> (first use in this function)
> > >>> 38 | #define FIBHEAPKEY_MIN LONG_MIN
> > >>> | ^~~~~~~~
> > >>> [snip]
> > >> What SHA1 of https://github.com/kalray/build-scripts are you using?
> > > I have executed the "source last.refs"
> >
> > I was referring to the SHA1 of the repo itself (build-scripts).
> >
> > `last.refs` is a symbolic link which can point to several releases,
> > depending on "when" you did the clone.
> >
> > I am asking this because we recently published new toolchains.
> >
> > I want to make sure which one you are trying to build.
>
> Unfortunately I deleted this repo a few minutes before you asked me ;-(
> But I remember that I cloned this repo two days ago.
> it should be: last.refs -> refs/4.11.0.refs
It should be my own environmental problem.
I reinstalled the system once and it has been able to compile normally ;-)
In the past few days, I have reviewed almost all the codes,
which is very meaningful for me to learn, thank you team.
>
> > >> We are building our toolchain on Ubuntu 18.04 / 20.04 and 22.04 without
> > >> issues, I don't understand why it does not work for you, although indeed
> > >> the error log you are having pops out on my search engine and seems to
> > >> be some well known issue.
> > > Yes, there are many answers on the web, but none of them solve this problem.
> > >
> > >> If the build-script does not work for you, you can still use the
> > >> pre-built toolchains generated by the GitHub automated actions:
> > >> https://github.com/kalray/build-scripts/releases/tag/v4.11.1 ("latest"
> > >> means 22.04)
> > > Thanks, this is the final solution ;-)
> > Good to see it helped :)
> >
> > Regards,
> >
> > --
> >
> > Yann
> >
> >
> >
> >
> >
>
>
> --
> Thanks,
> JeffXie
--
Thanks,
JeffXie
1 year, 11 months
audit rules to help watch for potential threat?
by Wieprecht, Karen M.
Steve, Audit team,
My colleagues and I were discussing ways we might better monitor for potential insider threat. We can easily see the commands our SAs run when they use sudo in front of the command, but if the sysadmin uses "sudo su -", then we don't have good visibility into the commands they perform while they are su'd unless there happens to be an audit rule monitoring the specific files/commands they are accessing/running.
We've talked about possible way to improve our visibility in this situation, but most of the options we came up with are easily thwarted and/or would cause the logs to blow up to the point that it's difficult to spot nefarious activity. Some options we considered included having splunk monitor the shell history files, and possibly enabling ps auditing.
Can you recommend any audit rules that would audit the interactive commands being issued by a sysadmin who is su'd as root without causing the logs to blow up?
Any assistance you can provide would be much appreciated.
Thank you,
Karen Wieprecht
The Johns Hopkins Applied Physics Laboratory
1 year, 11 months
[PATCH v3 1/2] bpf: restore the ebpf program ID for BPF_AUDIT_UNLOAD and PERF_BPF_EVENT_PROG_UNLOAD
by Paul Moore
When changing the ebpf program put() routines to support being called
from within IRQ context the program ID was reset to zero prior to
calling the perf event and audit UNLOAD record generators, which
resulted in problems as the ebpf program ID was bogus (always zero).
This patch addresses this problem by removing an unnecessary call to
bpf_prog_free_id() in __bpf_prog_offload_destroy() and adjusting
__bpf_prog_put() to only call bpf_prog_free_id() after audit and perf
have finished their bpf program unload tasks in
bpf_prog_put_deferred(). For the record, no one can determine, or
remember, why it was necessary to free the program ID, and remove it
from the IDR, prior to executing bpf_prog_put_deferred();
regardless, both Stanislav and Alexei agree that the approach in this
patch should be safe.
It is worth noting that when moving the bpf_prog_free_id() call, the
do_idr_lock parameter was forced to true as the ebpf devs determined
this was the correct as the do_idr_lock should always be true. The
do_idr_lock parameter will be removed in a follow-up patch, but it
was kept here to keep the patch small in an effort to ease any stable
backports.
I also modified the bpf_audit_prog() logic used to associate the
AUDIT_BPF record with other associated records, e.g. @ctx != NULL.
Instead of keying off the operation, it now keys off the execution
context, e.g. '!in_irg && !irqs_disabled()', which is much more
appropriate and should help better connect the UNLOAD operations with
the associated audit state (other audit records).
Cc: stable(a)vger.kernel.org
Fixes: d809e134be7a ("bpf: Prepare bpf_prog_put() to be called from irq context.")
Reported-by: Burn Alting <burn.alting(a)iinet.net.au>
Reported-by: Jiri Olsa <olsajiri(a)gmail.com>
Suggested-by: Stanislav Fomichev <sdf(a)google.com>
Suggested-by: Alexei Starovoitov <alexei.starovoitov(a)gmail.com>
Signed-off-by: Paul Moore <paul(a)paul-moore.com>
---
* v3
- abandon most of the changes in v2
- move bpf_prog_free_id() after the audit/perf unload hooks
- remove bpf_prog_free_id() from __bpf_prog_offload_destroy()
- added stable tag
* v2
- change subj
- add mention of the perf regression
- drop the dedicated program audit ID
- add the bpf_prog::valid_id flag, bpf_prog_get_id() getter
- convert prog ID users to new ID getter
* v1
- subj was: "bpf: restore the ebpf audit UNLOAD id field"
- initial draft
---
kernel/bpf/offload.c | 3 ---
kernel/bpf/syscall.c | 6 ++----
2 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c
index 13e4efc971e6..190d9f9dc987 100644
--- a/kernel/bpf/offload.c
+++ b/kernel/bpf/offload.c
@@ -216,9 +216,6 @@ static void __bpf_prog_offload_destroy(struct bpf_prog *prog)
if (offload->dev_state)
offload->offdev->ops->destroy(prog);
- /* Make sure BPF_PROG_GET_NEXT_ID can't find this dead program */
- bpf_prog_free_id(prog, true);
-
list_del_init(&offload->offloads);
kfree(offload);
prog->aux->offload = NULL;
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 64131f88c553..61bb19e81b9c 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1972,7 +1972,7 @@ static void bpf_audit_prog(const struct bpf_prog *prog, unsigned int op)
return;
if (audit_enabled == AUDIT_OFF)
return;
- if (op == BPF_AUDIT_LOAD)
+ if (!in_irq() && !irqs_disabled())
ctx = audit_context();
ab = audit_log_start(ctx, GFP_ATOMIC, AUDIT_BPF);
if (unlikely(!ab))
@@ -2067,6 +2067,7 @@ static void bpf_prog_put_deferred(struct work_struct *work)
prog = aux->prog;
perf_event_bpf_event(prog, PERF_BPF_EVENT_PROG_UNLOAD, 0);
bpf_audit_prog(prog, BPF_AUDIT_UNLOAD);
+ bpf_prog_free_id(prog, true);
__bpf_prog_put_noref(prog, true);
}
@@ -2075,9 +2076,6 @@ static void __bpf_prog_put(struct bpf_prog *prog, bool do_idr_lock)
struct bpf_prog_aux *aux = prog->aux;
if (atomic64_dec_and_test(&aux->refcnt)) {
- /* bpf_prog_free_id() must be called first */
- bpf_prog_free_id(prog, do_idr_lock);
-
if (in_irq() || irqs_disabled()) {
INIT_WORK(&aux->work, bpf_prog_put_deferred);
schedule_work(&aux->work);
--
2.39.0
1 year, 11 months
Identify whether the kernel version supports Path based exclusion
by Anurag Aggarwal
Hello All,
I need a method to identify whether the audid version a kernel is running
supports path based exclusions.
One option would be to use audit_add_rule_data to add a temporary path
based rule and check if it is successful, but this won't work when auditd
is running in immutable mode.
Any other way which does not require checking versions of Kernel or
Distribution?
--
Anurag Aggarwal
1 year, 11 months
New bug in Audit
by Ariel Silver
I found the following bug:
OS version = Red Hat Enterprise Linux release 8.6 (Ootpa)
Kernel version = 4.18.0-425.3.1.el8.x86_64
auditctl version = 3.0.7
Scenario 1:
When I load the configurations :
*auditctl -a always,exit -S all -F dir=/ -F perm=w -F success=1*
And run the command:
*cp /tmp/1 /tmp/2*
No new log is created in: /var/log/audit/audit.log
But the file is indeed copied.
Scenario 2:
When I load the configurations :
*auditctl -a always,exit -S all -F dir=/ -F perm=w -F success=0*
And run the command:
*cp /tmp/1 /tmp/2*
No new log is created in: /var/log/audit/audit.log
But the file is indeed copied.
Scenario 3:
When I load the configurations :
*auditctl -a always,exit -S all -F dir=/ -F perm=w*
And run the command:
*cp /tmp/1 /tmp/2*
Yes new log is created in: /var/log/audit/audit.log
File was indeed copied.
Conclusion:
Only when I don't use the -F success new logs are created.
Why is that?
Any alternative ?
1 year, 11 months
[PATCH v2] bpf: restore the ebpf program ID for BPF_AUDIT_UNLOAD and PERF_BPF_EVENT_PROG_UNLOAD
by Paul Moore
When changing the ebpf program put() routines to support being called
from within IRQ context the program ID was reset to zero prior to
calling the perf event and audit UNLOAD record generators, which
resulted in problems as the ebpf program ID was bogus (always zero).
This patch resolves this by adding a new flag, bpf_prog::valid_id, to
indicate when the bpf_prog_aux ID field is valid; it is set to true/1
in bpf_prog_alloc_id() and set to false/0 in bpf_prog_free_id(). In
order to help ensure that access to the bpf_prog_aux ID field takes
into account the new valid_id flag, the bpf_prog_aux ID field is
renamed to bpf_prog_aux::__id and a getter function,
bpf_prog_get_id(), was created and all users of bpf_prog_aux::id were
converted to the new caller. Exceptions to this include some of the
internal ebpf functions and the xdp trace points, although the latter
still take into account the valid_id flag.
I also modified the bpf_audit_prog() logic used to associate the
AUDIT_BPF record with other associated records, e.g. @ctx != NULL.
Instead of keying off the operation, it now keys off the execution
context, e.g. '!in_irg && !irqs_disabled()', which is much more
appropriate and should help better connect the UNLOAD operations with
the associated audit state (other audit records).
Fixes: d809e134be7a ("bpf: Prepare bpf_prog_put() to be called from irq context.")
Reported-by: Burn Alting <burn.alting(a)iinet.net.au>
Reported-by: Jiri Olsa <olsajiri(a)gmail.com>
Signed-off-by: Paul Moore <paul(a)paul-moore.com>
--
* v2
- change subj
- add mention of the perf regression
- drop the dedicated program audit ID
- add the bpf_prog::valid_id flag, bpf_prog_get_id() getter
- convert prog ID users to new ID getter
* v1
- subj was: "bpf: restore the ebpf audit UNLOAD id field"
- initial draft
---
drivers/net/netdevsim/bpf.c | 6 ++++--
include/linux/bpf.h | 11 +++++++++--
include/linux/bpf_verifier.h | 2 +-
include/trace/events/xdp.h | 4 ++--
kernel/bpf/arraymap.c | 2 +-
kernel/bpf/bpf_struct_ops.c | 2 +-
kernel/bpf/cgroup.c | 2 +-
kernel/bpf/core.c | 2 +-
kernel/bpf/cpumap.c | 2 +-
kernel/bpf/devmap.c | 2 +-
kernel/bpf/syscall.c | 27 +++++++++++++++------------
kernel/events/core.c | 6 +++++-
kernel/trace/bpf_trace.c | 2 +-
net/core/dev.c | 2 +-
net/core/filter.c | 3 ++-
net/core/rtnetlink.c | 2 +-
net/core/sock_map.c | 2 +-
net/ipv6/seg6_local.c | 3 ++-
net/sched/act_bpf.c | 2 +-
net/sched/cls_bpf.c | 2 +-
20 files changed, 52 insertions(+), 34 deletions(-)
diff --git a/drivers/net/netdevsim/bpf.c b/drivers/net/netdevsim/bpf.c
index 50854265864d..2795f03f5f34 100644
--- a/drivers/net/netdevsim/bpf.c
+++ b/drivers/net/netdevsim/bpf.c
@@ -109,7 +109,7 @@ nsim_bpf_offload(struct netdevsim *ns, struct bpf_prog *prog, bool oldprog)
"bad offload state, expected offload %sto be active",
oldprog ? "" : "not ");
ns->bpf_offloaded = prog;
- ns->bpf_offloaded_id = prog ? prog->aux->id : 0;
+ ns->bpf_offloaded_id = prog ? bpf_prog_get_id(prog) : 0;
nsim_prog_set_loaded(prog, true);
return 0;
@@ -221,6 +221,7 @@ static int nsim_bpf_create_prog(struct nsim_dev *nsim_dev,
struct nsim_bpf_bound_prog *state;
char name[16];
int ret;
+ u32 id;
state = kzalloc(sizeof(*state), GFP_KERNEL);
if (!state)
@@ -239,7 +240,8 @@ static int nsim_bpf_create_prog(struct nsim_dev *nsim_dev,
return ret;
}
- debugfs_create_u32("id", 0400, state->ddir, &prog->aux->id);
+ id = bpf_prog_get_id(prog);
+ debugfs_create_u32("id", 0400, state->ddir, &id);
debugfs_create_file("state", 0400, state->ddir,
&state->state, &nsim_bpf_string_fops);
debugfs_create_bool("loaded", 0400, state->ddir, &state->is_loaded);
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 9e7d46d16032..18e965bd7db9 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -1102,7 +1102,7 @@ struct bpf_prog_aux {
u32 max_pkt_offset;
u32 max_tp_access;
u32 stack_depth;
- u32 id;
+ u32 __id; /* access via bpf_prog_get_id() to check bpf_prog::valid_id */
u32 func_cnt; /* used by non-func prog as the number of func progs */
u32 func_idx; /* 0 for non-func prog, the index in func array for func prog */
u32 attach_btf_id; /* in-kernel BTF type id to attach to */
@@ -1197,7 +1197,8 @@ struct bpf_prog {
enforce_expected_attach_type:1, /* Enforce expected_attach_type checking at attach time */
call_get_stack:1, /* Do we call bpf_get_stack() or bpf_get_stackid() */
call_get_func_ip:1, /* Do we call get_func_ip() */
- tstamp_type_access:1; /* Accessed __sk_buff->tstamp_type */
+ tstamp_type_access:1, /* Accessed __sk_buff->tstamp_type */
+ valid_id:1; /* Is bpf_prog::aux::__id valid? */
enum bpf_prog_type type; /* Type of BPF program */
enum bpf_attach_type expected_attach_type; /* For some prog types */
u32 len; /* Number of filter blocks */
@@ -1688,6 +1689,12 @@ void bpf_prog_inc(struct bpf_prog *prog);
struct bpf_prog * __must_check bpf_prog_inc_not_zero(struct bpf_prog *prog);
void bpf_prog_put(struct bpf_prog *prog);
+static inline u32 bpf_prog_get_id(const struct bpf_prog *prog)
+{
+ if (WARN(!prog->valid_id, "Attempting to use an invalid eBPF program"))
+ return 0;
+ return prog->aux->__id;
+}
void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock);
void bpf_map_free_id(struct bpf_map *map, bool do_idr_lock);
diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h
index 9e1e6965f407..525c02cc12ea 100644
--- a/include/linux/bpf_verifier.h
+++ b/include/linux/bpf_verifier.h
@@ -604,7 +604,7 @@ static inline u64 bpf_trampoline_compute_key(const struct bpf_prog *tgt_prog,
struct btf *btf, u32 btf_id)
{
if (tgt_prog)
- return ((u64)tgt_prog->aux->id << 32) | btf_id;
+ return ((u64)bpf_prog_get_id(tgt_prog) << 32) | btf_id;
else
return ((u64)btf_obj_id(btf) << 32) | 0x80000000 | btf_id;
}
diff --git a/include/trace/events/xdp.h b/include/trace/events/xdp.h
index c40fc97f9417..a1c3048872ea 100644
--- a/include/trace/events/xdp.h
+++ b/include/trace/events/xdp.h
@@ -39,7 +39,7 @@ TRACE_EVENT(xdp_exception,
),
TP_fast_assign(
- __entry->prog_id = xdp->aux->id;
+ __entry->prog_id = (xdp->valid_id ? xdp->aux->__id : 0);
__entry->act = act;
__entry->ifindex = dev->ifindex;
),
@@ -120,7 +120,7 @@ DECLARE_EVENT_CLASS(xdp_redirect_template,
map_index = 0;
}
- __entry->prog_id = xdp->aux->id;
+ __entry->prog_id = (xdp->valid_id ? xdp->aux->__id : 0);
__entry->act = XDP_REDIRECT;
__entry->ifindex = dev->ifindex;
__entry->err = err;
diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
index 832b2659e96e..d19db5980b1b 100644
--- a/kernel/bpf/arraymap.c
+++ b/kernel/bpf/arraymap.c
@@ -905,7 +905,7 @@ static void prog_fd_array_put_ptr(void *ptr)
static u32 prog_fd_array_sys_lookup_elem(void *ptr)
{
- return ((struct bpf_prog *)ptr)->aux->id;
+ return bpf_prog_get_id((struct bpf_prog *)ptr);
}
/* decrement refcnt of all bpf_progs that are stored in this map */
diff --git a/kernel/bpf/bpf_struct_ops.c b/kernel/bpf/bpf_struct_ops.c
index 84b2d9dba79a..6c20e6cd9442 100644
--- a/kernel/bpf/bpf_struct_ops.c
+++ b/kernel/bpf/bpf_struct_ops.c
@@ -488,7 +488,7 @@ static int bpf_struct_ops_map_update_elem(struct bpf_map *map, void *key,
image += err;
/* put prog_id to udata */
- *(unsigned long *)(udata + moff) = prog->aux->id;
+ *(unsigned long *)(udata + moff) = bpf_prog_get_id(prog);
}
refcount_set(&kvalue->refcnt, 1);
diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index bf2fdb33fb31..4a8d26f1d5d1 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -1091,7 +1091,7 @@ static int __cgroup_bpf_query(struct cgroup *cgrp, const union bpf_attr *attr,
i = 0;
hlist_for_each_entry(pl, progs, node) {
prog = prog_list_prog(pl);
- id = prog->aux->id;
+ id = bpf_prog_get_id(prog);
if (copy_to_user(prog_ids + i, &id, sizeof(id)))
return -EFAULT;
if (++i == cnt)
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 25a54e04560e..ea3938ab6f5b 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -2293,7 +2293,7 @@ static bool bpf_prog_array_copy_core(struct bpf_prog_array *array,
for (item = array->items; item->prog; item++) {
if (item->prog == &dummy_bpf_prog.prog)
continue;
- prog_ids[i] = item->prog->aux->id;
+ prog_ids[i] = bpf_prog_get_id(item->prog);
if (++i == request_cnt) {
item++;
break;
diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
index b5ba34ddd4b6..3f3423d03aea 100644
--- a/kernel/bpf/cpumap.c
+++ b/kernel/bpf/cpumap.c
@@ -413,7 +413,7 @@ static int __cpu_map_load_bpf_program(struct bpf_cpu_map_entry *rcpu,
return -EINVAL;
}
- rcpu->value.bpf_prog.id = prog->aux->id;
+ rcpu->value.bpf_prog.id = bpf_prog_get_id(prog);
rcpu->prog = prog;
return 0;
diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
index f9a87dcc5535..d46309d4aa9e 100644
--- a/kernel/bpf/devmap.c
+++ b/kernel/bpf/devmap.c
@@ -868,7 +868,7 @@ static struct bpf_dtab_netdev *__dev_map_alloc_node(struct net *net,
dev->dtab = dtab;
if (prog) {
dev->xdp_prog = prog;
- dev->val.bpf_prog.id = prog->aux->id;
+ dev->val.bpf_prog.id = bpf_prog_get_id(prog);
} else {
dev->xdp_prog = NULL;
dev->val.bpf_prog.id = 0;
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 7b373a5e861f..9e862ef792cb 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1958,13 +1958,14 @@ static void bpf_audit_prog(const struct bpf_prog *prog, unsigned int op)
return;
if (audit_enabled == AUDIT_OFF)
return;
- if (op == BPF_AUDIT_LOAD)
+ if (!in_irq() && !irqs_disabled())
ctx = audit_context();
ab = audit_log_start(ctx, GFP_ATOMIC, AUDIT_BPF);
if (unlikely(!ab))
return;
+ /* log the id regardless of bpf_prog::valid_id */
audit_log_format(ab, "prog-id=%u op=%s",
- prog->aux->id, bpf_audit_str[op]);
+ prog->aux->__id, bpf_audit_str[op]);
audit_log_end(ab);
}
@@ -1975,8 +1976,10 @@ static int bpf_prog_alloc_id(struct bpf_prog *prog)
idr_preload(GFP_KERNEL);
spin_lock_bh(&prog_idr_lock);
id = idr_alloc_cyclic(&prog_idr, prog, 1, INT_MAX, GFP_ATOMIC);
- if (id > 0)
- prog->aux->id = id;
+ if (id > 0) {
+ prog->aux->__id = id;
+ prog->valid_id = true;
+ }
spin_unlock_bh(&prog_idr_lock);
idr_preload_end();
@@ -1996,7 +1999,7 @@ void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock)
* disappears - even if someone grabs an fd to them they are unusable,
* simply waiting for refcnt to drop to be freed.
*/
- if (!prog->aux->id)
+ if (!prog->valid_id)
return;
if (do_idr_lock)
@@ -2004,8 +2007,8 @@ void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock)
else
__acquire(&prog_idr_lock);
- idr_remove(&prog_idr, prog->aux->id);
- prog->aux->id = 0;
+ idr_remove(&prog_idr, prog->aux->__id);
+ prog->valid_id = false;
if (do_idr_lock)
spin_unlock_irqrestore(&prog_idr_lock, flags);
@@ -2154,7 +2157,7 @@ static void bpf_prog_show_fdinfo(struct seq_file *m, struct file *filp)
prog->jited,
prog_tag,
prog->pages * 1ULL << PAGE_SHIFT,
- prog->aux->id,
+ bpf_prog_get_id(prog),
stats.nsecs,
stats.cnt,
stats.misses,
@@ -2786,7 +2789,7 @@ static void bpf_link_show_fdinfo(struct seq_file *m, struct file *filp)
bpf_link_type_strs[link->type],
link->id,
prog_tag,
- prog->aux->id);
+ bpf_prog_get_id(prog));
if (link->ops->show_fdinfo)
link->ops->show_fdinfo(link, m);
}
@@ -3914,7 +3917,7 @@ static int bpf_prog_get_info_by_fd(struct file *file,
return -EFAULT;
info.type = prog->type;
- info.id = prog->aux->id;
+ info.id = bpf_prog_get_id(prog);
info.load_time = prog->aux->load_time;
info.created_by_uid = from_kuid_munged(current_user_ns(),
prog->aux->user->uid);
@@ -4261,7 +4264,7 @@ static int bpf_link_get_info_by_fd(struct file *file,
info.type = link->type;
info.id = link->id;
- info.prog_id = link->prog->aux->id;
+ info.prog_id = bpf_prog_get_id(link->prog);
if (link->ops->fill_link_info) {
err = link->ops->fill_link_info(link, &info);
@@ -4426,7 +4429,7 @@ static int bpf_task_fd_query(const union bpf_attr *attr,
struct bpf_raw_event_map *btp = raw_tp->btp;
err = bpf_task_fd_query_copy(attr, uattr,
- raw_tp->link.prog->aux->id,
+ bpf_prog_get_id(raw_tp->link.prog),
BPF_FD_TYPE_RAW_TRACEPOINT,
btp->tp->name, 0, 0);
goto put_file;
diff --git a/kernel/events/core.c b/kernel/events/core.c
index aefc1e08e015..c24e897d27f1 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -9001,7 +9001,11 @@ void perf_event_bpf_event(struct bpf_prog *prog,
},
.type = type,
.flags = flags,
- .id = prog->aux->id,
+ /*
+ * don't use bpf_prog_get_id() as the id may be marked
+ * invalid on PERF_BPF_EVENT_PROG_UNLOAD events
+ */
+ .id = prog->aux->__id,
},
};
diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index 49fb9ec8366d..7cd0eb83b137 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -2344,7 +2344,7 @@ int bpf_get_perf_event_info(const struct perf_event *event, u32 *prog_id,
if (prog->type == BPF_PROG_TYPE_PERF_EVENT)
return -EOPNOTSUPP;
- *prog_id = prog->aux->id;
+ *prog_id = bpf_prog_get_id(prog);
flags = event->tp_event->flags;
is_tracepoint = flags & TRACE_EVENT_FL_TRACEPOINT;
is_syscall_tp = is_syscall_trace_event(event->tp_event);
diff --git a/net/core/dev.c b/net/core/dev.c
index fa53830d0683..0d39ef22cf4b 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9068,7 +9068,7 @@ u32 dev_xdp_prog_id(struct net_device *dev, enum bpf_xdp_mode mode)
{
struct bpf_prog *prog = dev_xdp_prog(dev, mode);
- return prog ? prog->aux->id : 0;
+ return prog ? bpf_prog_get_id(prog) : 0;
}
static void dev_xdp_set_link(struct net_device *dev, enum bpf_xdp_mode mode,
diff --git a/net/core/filter.c b/net/core/filter.c
index bb0136e7a8e4..282ccfe34ced 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -8729,7 +8729,8 @@ void bpf_warn_invalid_xdp_action(struct net_device *dev, struct bpf_prog *prog,
pr_warn_once("%s XDP return value %u on prog %s (id %d) dev %s, expect packet loss!\n",
act > act_max ? "Illegal" : "Driver unsupported",
- act, prog->aux->name, prog->aux->id, dev ? dev->name : "N/A");
+ act, prog->aux->name, bpf_prog_get_id(prog),
+ dev ? dev->name : "N/A");
}
EXPORT_SYMBOL_GPL(bpf_warn_invalid_xdp_action);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 74864dc46a7e..1f7e36909541 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -1453,7 +1453,7 @@ static u32 rtnl_xdp_prog_skb(struct net_device *dev)
generic_xdp_prog = rtnl_dereference(dev->xdp_prog);
if (!generic_xdp_prog)
return 0;
- return generic_xdp_prog->aux->id;
+ return bpf_prog_get_id(generic_xdp_prog);
}
static u32 rtnl_xdp_prog_drv(struct net_device *dev)
diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index a660baedd9e7..550ec6cb3aee 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -1518,7 +1518,7 @@ int sock_map_bpf_prog_query(const union bpf_attr *attr,
/* we do not hold the refcnt, the bpf prog may be released
* asynchronously and the id would be set to 0.
*/
- id = data_race(prog->aux->id);
+ id = data_race(bpf_prog_get_id(prog));
if (id == 0)
prog_cnt = 0;
diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c
index 8370726ae7bf..440ce3aba802 100644
--- a/net/ipv6/seg6_local.c
+++ b/net/ipv6/seg6_local.c
@@ -1543,7 +1543,8 @@ static int put_nla_bpf(struct sk_buff *skb, struct seg6_local_lwt *slwt)
if (!nest)
return -EMSGSIZE;
- if (nla_put_u32(skb, SEG6_LOCAL_BPF_PROG, slwt->bpf.prog->aux->id))
+ if (nla_put_u32(skb, SEG6_LOCAL_BPF_PROG,
+ bpf_prog_get_id(slwt->bpf.prog)))
return -EMSGSIZE;
if (slwt->bpf.name &&
diff --git a/net/sched/act_bpf.c b/net/sched/act_bpf.c
index b79eee44e24e..604a29e482b0 100644
--- a/net/sched/act_bpf.c
+++ b/net/sched/act_bpf.c
@@ -121,7 +121,7 @@ static int tcf_bpf_dump_ebpf_info(const struct tcf_bpf *prog,
nla_put_string(skb, TCA_ACT_BPF_NAME, prog->bpf_name))
return -EMSGSIZE;
- if (nla_put_u32(skb, TCA_ACT_BPF_ID, prog->filter->aux->id))
+ if (nla_put_u32(skb, TCA_ACT_BPF_ID, bpf_prog_get_id(prog->filter)))
return -EMSGSIZE;
nla = nla_reserve(skb, TCA_ACT_BPF_TAG, sizeof(prog->filter->tag));
diff --git a/net/sched/cls_bpf.c b/net/sched/cls_bpf.c
index bc317b3eac12..eb5ac6be589e 100644
--- a/net/sched/cls_bpf.c
+++ b/net/sched/cls_bpf.c
@@ -565,7 +565,7 @@ static int cls_bpf_dump_ebpf_info(const struct cls_bpf_prog *prog,
nla_put_string(skb, TCA_BPF_NAME, prog->bpf_name))
return -EMSGSIZE;
- if (nla_put_u32(skb, TCA_BPF_ID, prog->filter->aux->id))
+ if (nla_put_u32(skb, TCA_BPF_ID, bpf_prog_get_id(prog->filter)))
return -EMSGSIZE;
nla = nla_reserve(skb, TCA_BPF_TAG, sizeof(prog->filter->tag));
--
2.39.0
1 year, 11 months
