Skip to content

Commit

Permalink
Revert "ANDROID: security,perf: Allow further restriction of perf_eve…
Browse files Browse the repository at this point in the history
…nt_open"

Unfork Android.

This reverts commit 8e5e42d5ae20f0324170d01ccf374a1571e82d9b.

Perf_event_paranoid=3 is no longer needed on Android. Access control
of perf events is now done by selinux. See:
https://patchwork.kernel.org/patch/11185793/

Bug: 120445712
Bug: 137092007
Signed-off-by: Jeff Vander Stoep <[email protected]>
Change-Id: Iba493424174b30baff460caaa25a54a472c87bd4
  • Loading branch information
jeffvanderstoep authored and duhansysl committed Sep 5, 2024
1 parent f6d371a commit a8d0c73
Show file tree
Hide file tree
Showing 6 changed files with 2 additions and 26 deletions.
4 changes: 1 addition & 3 deletions Documentation/sysctl/kernel.txt
Original file line number Diff line number Diff line change
Expand Up @@ -660,14 +660,12 @@ allowed to execute.
perf_event_paranoid:

Controls use of the performance events system by unprivileged
users (without CAP_SYS_ADMIN). The default value is 3 if
CONFIG_SECURITY_PERF_EVENTS_RESTRICT is set, or 2 otherwise.
users (without CAP_SYS_ADMIN). The default value is 2.

-1: Allow use of (almost) all events by all users
>=0: Disallow raw tracepoint access by users without CAP_IOC_LOCK
>=1: Disallow CPU event access by users without CAP_SYS_ADMIN
>=2: Disallow kernel profiling by users without CAP_SYS_ADMIN
>=3: Disallow all event access by users without CAP_SYS_ADMIN

==============================================================

Expand Down
1 change: 0 additions & 1 deletion arch/arm64/configs/exynos9810_defconfig
Original file line number Diff line number Diff line change
Expand Up @@ -5725,7 +5725,6 @@ CONFIG_KEYS=y
# CONFIG_ENCRYPTED_KEYS is not set
# CONFIG_KEY_DH_OPERATIONS is not set
# CONFIG_SECURITY_DMESG_RESTRICT is not set
CONFIG_SECURITY_PERF_EVENTS_RESTRICT=y
CONFIG_SECURITY=y
# CONFIG_SECURITYFS is not set
CONFIG_SECURITY_NETWORK=y
Expand Down
1 change: 1 addition & 0 deletions arch/x86/configs/x86_64_cuttlefish_defconfig
Original file line number Diff line number Diff line change
Expand Up @@ -487,6 +487,7 @@ CONFIG_IO_DELAY_NONE=y
CONFIG_DEBUG_BOOT_PARAMS=y
CONFIG_OPTIMIZE_INLINING=y
CONFIG_SECURITY_PERF_EVENTS_RESTRICT=y
CONFIG_UNWINDER_FRAME_POINTER=y
CONFIG_SECURITY=y
CONFIG_SECURITY_NETWORK=y
CONFIG_SECURITY_PATH=y
Expand Down
5 changes: 0 additions & 5 deletions include/linux/perf_event.h
Original file line number Diff line number Diff line change
Expand Up @@ -1185,11 +1185,6 @@ int perf_event_max_stack_handler(struct ctl_table *table, int write,
#define PERF_SECURITY_KERNEL 2
#define PERF_SECURITY_TRACEPOINT 3

static inline bool perf_paranoid_any(void)
{
return sysctl_perf_event_paranoid > 2;
}

static inline int perf_is_paranoid(void)
{
return sysctl_perf_event_paranoid > -1;
Expand Down
8 changes: 0 additions & 8 deletions kernel/events/core.c
Original file line number Diff line number Diff line change
Expand Up @@ -395,13 +395,8 @@ static struct srcu_struct pmus_srcu;
* 0 - disallow raw tracepoint access for unpriv
* 1 - disallow cpu events for unpriv
* 2 - disallow kernel profiling for unpriv
* 3 - disallow all unpriv perf event use
*/
#ifdef CONFIG_SECURITY_PERF_EVENTS_RESTRICT
int sysctl_perf_event_paranoid __read_mostly = 3;
#else
int sysctl_perf_event_paranoid __read_mostly = 2;
#endif

/* Minimum for 512 kiB + 1 user control page */
int sysctl_perf_event_mlock __read_mostly = 512 + (PAGE_SIZE / 1024); /* 'free' kiB per user */
Expand Down Expand Up @@ -9760,9 +9755,6 @@ SYSCALL_DEFINE5(perf_event_open,
if (flags & ~PERF_FLAG_ALL)
return -EINVAL;

if (perf_paranoid_any() && !capable(CAP_SYS_ADMIN))
return -EACCES;

/* Do we allow access to perf_event_open(2) ? */
err = security_perf_event_open(&attr, PERF_SECURITY_OPEN);
if (err)
Expand Down
9 changes: 0 additions & 9 deletions security/Kconfig
Original file line number Diff line number Diff line change
Expand Up @@ -18,15 +18,6 @@ config SECURITY_DMESG_RESTRICT

If you are unsure how to answer this question, answer N.

config SECURITY_PERF_EVENTS_RESTRICT
bool "Restrict unprivileged use of performance events"
depends on PERF_EVENTS
help
If you say Y here, the kernel.perf_event_paranoid sysctl
will be set to 3 by default, and no unprivileged use of the
perf_event_open syscall will be permitted unless it is
changed.

config SECURITY
bool "Enable different security models"
depends on SYSFS
Expand Down

0 comments on commit a8d0c73

Please sign in to comment.