The patch below does not apply to the 5.10-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
>From 85b6d24646e4125c591639841169baa98a2da503 Mon Sep 17 00:00:00 2001
From: Alexander Mikhalitsyn <alexander.mikhalitsyn(a)virtuozzo.com>
Date: Fri, 19 Nov 2021 16:43:21 -0800
Subject: [PATCH] shm: extend forced shm destroy to support objects from
several IPC nses
Currently, the exit_shm() function not designed to work properly when
task->sysvshm.shm_clist holds shm objects from different IPC namespaces.
This is a real pain when sysctl kernel.shm_rmid_forced = 1, because it
leads to use-after-free (reproducer exists).
This is an attempt to fix the problem by extending exit_shm mechanism to
handle shm's destroy from several IPC ns'es.
To achieve that we do several things:
1. add a namespace (non-refcounted) pointer to the struct shmid_kernel
2. during new shm object creation (newseg()/shmget syscall) we
initialize this pointer by current task IPC ns
3. exit_shm() fully reworked such that it traverses over all shp's in
task->sysvshm.shm_clist and gets IPC namespace not from current task
as it was before but from shp's object itself, then call
shm_destroy(shp, ns).
Note: We need to be really careful here, because as it was said before
(1), our pointer to IPC ns non-refcnt'ed. To be on the safe side we
using special helper get_ipc_ns_not_zero() which allows to get IPC ns
refcounter only if IPC ns not in the "state of destruction".
Q/A
Q: Why can we access shp->ns memory using non-refcounted pointer?
A: Because shp object lifetime is always shorther than IPC namespace
lifetime, so, if we get shp object from the task->sysvshm.shm_clist
while holding task_lock(task) nobody can steal our namespace.
Q: Does this patch change semantics of unshare/setns/clone syscalls?
A: No. It's just fixes non-covered case when process may leave IPC
namespace without getting task->sysvshm.shm_clist list cleaned up.
Link: https://lkml.kernel.org/r/67bb03e5-f79c-1815-e2bf-949c67047418@colorfullife…
Link: https://lkml.kernel.org/r/20211109151501.4921-1-manfred@colorfullife.com
Fixes: ab602f79915 ("shm: make exit_shm work proportional to task activity")
Co-developed-by: Manfred Spraul <manfred(a)colorfullife.com>
Signed-off-by: Manfred Spraul <manfred(a)colorfullife.com>
Signed-off-by: Alexander Mikhalitsyn <alexander.mikhalitsyn(a)virtuozzo.com>
Cc: "Eric W. Biederman" <ebiederm(a)xmission.com>
Cc: Davidlohr Bueso <dave(a)stgolabs.net>
Cc: Greg KH <gregkh(a)linuxfoundation.org>
Cc: Andrei Vagin <avagin(a)gmail.com>
Cc: Pavel Tikhomirov <ptikhomirov(a)virtuozzo.com>
Cc: Vasily Averin <vvs(a)virtuozzo.com>
Cc: <stable(a)vger.kernel.org>
Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org>
diff --git a/include/linux/ipc_namespace.h b/include/linux/ipc_namespace.h
index 05e22770af51..b75395ec8d52 100644
--- a/include/linux/ipc_namespace.h
+++ b/include/linux/ipc_namespace.h
@@ -131,6 +131,16 @@ static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns)
return ns;
}
+static inline struct ipc_namespace *get_ipc_ns_not_zero(struct ipc_namespace *ns)
+{
+ if (ns) {
+ if (refcount_inc_not_zero(&ns->ns.count))
+ return ns;
+ }
+
+ return NULL;
+}
+
extern void put_ipc_ns(struct ipc_namespace *ns);
#else
static inline struct ipc_namespace *copy_ipcs(unsigned long flags,
@@ -147,6 +157,11 @@ static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns)
return ns;
}
+static inline struct ipc_namespace *get_ipc_ns_not_zero(struct ipc_namespace *ns)
+{
+ return ns;
+}
+
static inline void put_ipc_ns(struct ipc_namespace *ns)
{
}
diff --git a/include/linux/sched/task.h b/include/linux/sched/task.h
index ba88a6987400..058d7f371e25 100644
--- a/include/linux/sched/task.h
+++ b/include/linux/sched/task.h
@@ -158,7 +158,7 @@ static inline struct vm_struct *task_stack_vm_area(const struct task_struct *t)
* Protects ->fs, ->files, ->mm, ->group_info, ->comm, keyring
* subscriptions and synchronises with wait4(). Also used in procfs. Also
* pins the final release of task.io_context. Also protects ->cpuset and
- * ->cgroup.subsys[]. And ->vfork_done.
+ * ->cgroup.subsys[]. And ->vfork_done. And ->sysvshm.shm_clist.
*
* Nests both inside and outside of read_lock(&tasklist_lock).
* It must not be nested with write_lock_irq(&tasklist_lock),
diff --git a/ipc/shm.c b/ipc/shm.c
index 4942bdd65748..b3048ebd5c31 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -62,9 +62,18 @@ struct shmid_kernel /* private to the kernel */
struct pid *shm_lprid;
struct ucounts *mlock_ucounts;
- /* The task created the shm object. NULL if the task is dead. */
+ /*
+ * The task created the shm object, for
+ * task_lock(shp->shm_creator)
+ */
struct task_struct *shm_creator;
- struct list_head shm_clist; /* list by creator */
+
+ /*
+ * List by creator. task_lock(->shm_creator) required for read/write.
+ * If list_empty(), then the creator is dead already.
+ */
+ struct list_head shm_clist;
+ struct ipc_namespace *ns;
} __randomize_layout;
/* shm_mode upper byte flags */
@@ -115,6 +124,7 @@ static void do_shm_rmid(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
struct shmid_kernel *shp;
shp = container_of(ipcp, struct shmid_kernel, shm_perm);
+ WARN_ON(ns != shp->ns);
if (shp->shm_nattch) {
shp->shm_perm.mode |= SHM_DEST;
@@ -225,10 +235,43 @@ static void shm_rcu_free(struct rcu_head *head)
kfree(shp);
}
-static inline void shm_rmid(struct ipc_namespace *ns, struct shmid_kernel *s)
+/*
+ * It has to be called with shp locked.
+ * It must be called before ipc_rmid()
+ */
+static inline void shm_clist_rm(struct shmid_kernel *shp)
{
- list_del(&s->shm_clist);
- ipc_rmid(&shm_ids(ns), &s->shm_perm);
+ struct task_struct *creator;
+
+ /* ensure that shm_creator does not disappear */
+ rcu_read_lock();
+
+ /*
+ * A concurrent exit_shm may do a list_del_init() as well.
+ * Just do nothing if exit_shm already did the work
+ */
+ if (!list_empty(&shp->shm_clist)) {
+ /*
+ * shp->shm_creator is guaranteed to be valid *only*
+ * if shp->shm_clist is not empty.
+ */
+ creator = shp->shm_creator;
+
+ task_lock(creator);
+ /*
+ * list_del_init() is a nop if the entry was already removed
+ * from the list.
+ */
+ list_del_init(&shp->shm_clist);
+ task_unlock(creator);
+ }
+ rcu_read_unlock();
+}
+
+static inline void shm_rmid(struct shmid_kernel *s)
+{
+ shm_clist_rm(s);
+ ipc_rmid(&shm_ids(s->ns), &s->shm_perm);
}
@@ -283,7 +326,7 @@ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
shm_file = shp->shm_file;
shp->shm_file = NULL;
ns->shm_tot -= (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT;
- shm_rmid(ns, shp);
+ shm_rmid(shp);
shm_unlock(shp);
if (!is_file_hugepages(shm_file))
shmem_lock(shm_file, 0, shp->mlock_ucounts);
@@ -303,10 +346,10 @@ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
*
* 2) sysctl kernel.shm_rmid_forced is set to 1.
*/
-static bool shm_may_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
+static bool shm_may_destroy(struct shmid_kernel *shp)
{
return (shp->shm_nattch == 0) &&
- (ns->shm_rmid_forced ||
+ (shp->ns->shm_rmid_forced ||
(shp->shm_perm.mode & SHM_DEST));
}
@@ -337,7 +380,7 @@ static void shm_close(struct vm_area_struct *vma)
ipc_update_pid(&shp->shm_lprid, task_tgid(current));
shp->shm_dtim = ktime_get_real_seconds();
shp->shm_nattch--;
- if (shm_may_destroy(ns, shp))
+ if (shm_may_destroy(shp))
shm_destroy(ns, shp);
else
shm_unlock(shp);
@@ -358,10 +401,10 @@ static int shm_try_destroy_orphaned(int id, void *p, void *data)
*
* As shp->* are changed under rwsem, it's safe to skip shp locking.
*/
- if (shp->shm_creator != NULL)
+ if (!list_empty(&shp->shm_clist))
return 0;
- if (shm_may_destroy(ns, shp)) {
+ if (shm_may_destroy(shp)) {
shm_lock_by_ptr(shp);
shm_destroy(ns, shp);
}
@@ -379,48 +422,97 @@ void shm_destroy_orphaned(struct ipc_namespace *ns)
/* Locking assumes this will only be called with task == current */
void exit_shm(struct task_struct *task)
{
- struct ipc_namespace *ns = task->nsproxy->ipc_ns;
- struct shmid_kernel *shp, *n;
+ for (;;) {
+ struct shmid_kernel *shp;
+ struct ipc_namespace *ns;
- if (list_empty(&task->sysvshm.shm_clist))
- return;
+ task_lock(task);
+
+ if (list_empty(&task->sysvshm.shm_clist)) {
+ task_unlock(task);
+ break;
+ }
+
+ shp = list_first_entry(&task->sysvshm.shm_clist, struct shmid_kernel,
+ shm_clist);
- /*
- * If kernel.shm_rmid_forced is not set then only keep track of
- * which shmids are orphaned, so that a later set of the sysctl
- * can clean them up.
- */
- if (!ns->shm_rmid_forced) {
- down_read(&shm_ids(ns).rwsem);
- list_for_each_entry(shp, &task->sysvshm.shm_clist, shm_clist)
- shp->shm_creator = NULL;
/*
- * Only under read lock but we are only called on current
- * so no entry on the list will be shared.
+ * 1) Get pointer to the ipc namespace. It is worth to say
+ * that this pointer is guaranteed to be valid because
+ * shp lifetime is always shorter than namespace lifetime
+ * in which shp lives.
+ * We taken task_lock it means that shp won't be freed.
*/
- list_del(&task->sysvshm.shm_clist);
- up_read(&shm_ids(ns).rwsem);
- return;
- }
+ ns = shp->ns;
- /*
- * Destroy all already created segments, that were not yet mapped,
- * and mark any mapped as orphan to cover the sysctl toggling.
- * Destroy is skipped if shm_may_destroy() returns false.
- */
- down_write(&shm_ids(ns).rwsem);
- list_for_each_entry_safe(shp, n, &task->sysvshm.shm_clist, shm_clist) {
- shp->shm_creator = NULL;
+ /*
+ * 2) If kernel.shm_rmid_forced is not set then only keep track of
+ * which shmids are orphaned, so that a later set of the sysctl
+ * can clean them up.
+ */
+ if (!ns->shm_rmid_forced)
+ goto unlink_continue;
- if (shm_may_destroy(ns, shp)) {
- shm_lock_by_ptr(shp);
- shm_destroy(ns, shp);
+ /*
+ * 3) get a reference to the namespace.
+ * The refcount could be already 0. If it is 0, then
+ * the shm objects will be free by free_ipc_work().
+ */
+ ns = get_ipc_ns_not_zero(ns);
+ if (!ns) {
+unlink_continue:
+ list_del_init(&shp->shm_clist);
+ task_unlock(task);
+ continue;
}
- }
- /* Remove the list head from any segments still attached. */
- list_del(&task->sysvshm.shm_clist);
- up_write(&shm_ids(ns).rwsem);
+ /*
+ * 4) get a reference to shp.
+ * This cannot fail: shm_clist_rm() is called before
+ * ipc_rmid(), thus the refcount cannot be 0.
+ */
+ WARN_ON(!ipc_rcu_getref(&shp->shm_perm));
+
+ /*
+ * 5) unlink the shm segment from the list of segments
+ * created by current.
+ * This must be done last. After unlinking,
+ * only the refcounts obtained above prevent IPC_RMID
+ * from destroying the segment or the namespace.
+ */
+ list_del_init(&shp->shm_clist);
+
+ task_unlock(task);
+
+ /*
+ * 6) we have all references
+ * Thus lock & if needed destroy shp.
+ */
+ down_write(&shm_ids(ns).rwsem);
+ shm_lock_by_ptr(shp);
+ /*
+ * rcu_read_lock was implicitly taken in shm_lock_by_ptr, it's
+ * safe to call ipc_rcu_putref here
+ */
+ ipc_rcu_putref(&shp->shm_perm, shm_rcu_free);
+
+ if (ipc_valid_object(&shp->shm_perm)) {
+ if (shm_may_destroy(shp))
+ shm_destroy(ns, shp);
+ else
+ shm_unlock(shp);
+ } else {
+ /*
+ * Someone else deleted the shp from namespace
+ * idr/kht while we have waited.
+ * Just unlock and continue.
+ */
+ shm_unlock(shp);
+ }
+
+ up_write(&shm_ids(ns).rwsem);
+ put_ipc_ns(ns); /* paired with get_ipc_ns_not_zero */
+ }
}
static vm_fault_t shm_fault(struct vm_fault *vmf)
@@ -676,7 +768,11 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
if (error < 0)
goto no_id;
+ shp->ns = ns;
+
+ task_lock(current);
list_add(&shp->shm_clist, ¤t->sysvshm.shm_clist);
+ task_unlock(current);
/*
* shmid gets reported as "inode#" in /proc/pid/maps.
@@ -1567,7 +1663,8 @@ long do_shmat(int shmid, char __user *shmaddr, int shmflg,
down_write(&shm_ids(ns).rwsem);
shp = shm_lock(ns, shmid);
shp->shm_nattch--;
- if (shm_may_destroy(ns, shp))
+
+ if (shm_may_destroy(shp))
shm_destroy(ns, shp);
else
shm_unlock(shp);
Hi
maybe we should add Fixes- and Stable-Tag to get it fixed in 5.16 and previous versions
i would suggest the commit introducing the warning shown when the skb->dev is missing
Fixes: 9b52e3f267a6 ("flow_dissector: handle no-skb use case")
Cc: stable(a)vger.kernel.org
Hello Greg, Sasha,
this series for 5.4-stable backports all the fixes (and their
dependencies) for Armada 3720 PCIe driver.
These include:
- fixes (and their dependencies) for pci-aardvark controller
- fixes for pci-bridge-emul
- fixes (and their dependencies) for pinctrl-armada-37xx driver
- device-tree fixes
Marek
Grzegorz Jaszczyk (1):
PCI: aardvark: Fix big endian support
Marek Behún (4):
PCI: aardvark: Improve link training
PCI: aardvark: Deduplicate code in advk_pcie_rd_conf()
pinctrl: armada-37xx: Correct PWM pins definitions
arm64: dts: marvell: armada-37xx: Set pcie_reset_pin to gpio function
Pali Rohár (15):
PCI: aardvark: Train link immediately after enabling training
PCI: aardvark: Issue PERST via GPIO
PCI: aardvark: Replace custom macros by standard linux/pci_regs.h
macros
PCI: aardvark: Don't touch PCIe registers if no card connected
PCI: aardvark: Fix compilation on s390
PCI: aardvark: Move PCIe reset card code to advk_pcie_train_link()
PCI: aardvark: Update comment about disabling link training
PCI: aardvark: Configure PCIe resources from 'ranges' DT property
PCI: aardvark: Fix PCIe Max Payload Size setting
PCI: aardvark: Implement re-issuing config requests on CRS response
PCI: aardvark: Simplify initialization of rootcap on virtual bridge
PCI: aardvark: Fix link training
PCI: aardvark: Fix support for bus mastering and PCI_COMMAND on
emulated bridge
PCI: aardvark: Set PCI Bridge Class Code to PCI Bridge
PCI: aardvark: Fix support for PCI_BRIDGE_CTL_BUS_RESET on emulated
bridge
Remi Pommarel (1):
PCI: aardvark: Wait for endpoint to be ready before training link
Russell King (1):
PCI: pci-bridge-emul: Fix array overruns, improve safety
.../pinctrl/marvell,armada-37xx-pinctrl.txt | 8 +-
.../arm64/boot/dts/marvell/armada-3720-db.dts | 3 +
.../dts/marvell/armada-3720-espressobin.dts | 1 +
.../dts/marvell/armada-3720-turris-mox.dts | 4 -
arch/arm64/boot/dts/marvell/armada-37xx.dtsi | 2 +-
drivers/pci/controller/pci-aardvark.c | 576 ++++++++++++++----
drivers/pci/pci-bridge-emul.c | 11 +-
drivers/pinctrl/mvebu/pinctrl-armada-37xx.c | 17 +-
8 files changed, 489 insertions(+), 133 deletions(-)
--
2.32.0
The patch below does not apply to the 5.10-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
>From 353050be4c19e102178ccc05988101887c25ae53 Mon Sep 17 00:00:00 2001
From: Daniel Borkmann <daniel(a)iogearbox.net>
Date: Tue, 9 Nov 2021 18:48:08 +0000
Subject: [PATCH] bpf: Fix toctou on read-only map's constant scalar tracking
Commit a23740ec43ba ("bpf: Track contents of read-only maps as scalars") is
checking whether maps are read-only both from BPF program side and user space
side, and then, given their content is constant, reading out their data via
map->ops->map_direct_value_addr() which is then subsequently used as known
scalar value for the register, that is, it is marked as __mark_reg_known()
with the read value at verification time. Before a23740ec43ba, the register
content was marked as an unknown scalar so the verifier could not make any
assumptions about the map content.
The current implementation however is prone to a TOCTOU race, meaning, the
value read as known scalar for the register is not guaranteed to be exactly
the same at a later point when the program is executed, and as such, the
prior made assumptions of the verifier with regards to the program will be
invalid which can cause issues such as OOB access, etc.
While the BPF_F_RDONLY_PROG map flag is always fixed and required to be
specified at map creation time, the map->frozen property is initially set to
false for the map given the map value needs to be populated, e.g. for global
data sections. Once complete, the loader "freezes" the map from user space
such that no subsequent updates/deletes are possible anymore. For the rest
of the lifetime of the map, this freeze one-time trigger cannot be undone
anymore after a successful BPF_MAP_FREEZE cmd return. Meaning, any new BPF_*
cmd calls which would update/delete map entries will be rejected with -EPERM
since map_get_sys_perms() removes the FMODE_CAN_WRITE permission. This also
means that pending update/delete map entries must still complete before this
guarantee is given. This corner case is not an issue for loaders since they
create and prepare such program private map in successive steps.
However, a malicious user is able to trigger this TOCTOU race in two different
ways: i) via userfaultfd, and ii) via batched updates. For i) userfaultfd is
used to expand the competition interval, so that map_update_elem() can modify
the contents of the map after map_freeze() and bpf_prog_load() were executed.
This works, because userfaultfd halts the parallel thread which triggered a
map_update_elem() at the time where we copy key/value from the user buffer and
this already passed the FMODE_CAN_WRITE capability test given at that time the
map was not "frozen". Then, the main thread performs the map_freeze() and
bpf_prog_load(), and once that had completed successfully, the other thread
is woken up to complete the pending map_update_elem() which then changes the
map content. For ii) the idea of the batched update is similar, meaning, when
there are a large number of updates to be processed, it can increase the
competition interval between the two. It is therefore possible in practice to
modify the contents of the map after executing map_freeze() and bpf_prog_load().
One way to fix both i) and ii) at the same time is to expand the use of the
map's map->writecnt. The latter was introduced in fc9702273e2e ("bpf: Add mmap()
support for BPF_MAP_TYPE_ARRAY") and further refined in 1f6cb19be2e2 ("bpf:
Prevent re-mmap()'ing BPF map as writable for initially r/o mapping") with
the rationale to make a writable mmap()'ing of a map mutually exclusive with
read-only freezing. The counter indicates writable mmap() mappings and then
prevents/fails the freeze operation. Its semantics can be expanded beyond
just mmap() by generally indicating ongoing write phases. This would essentially
span any parallel regular and batched flavor of update/delete operation and
then also have map_freeze() fail with -EBUSY. For the check_mem_access() in
the verifier we expand upon the bpf_map_is_rdonly() check ensuring that all
last pending writes have completed via bpf_map_write_active() test. Once the
map->frozen is set and bpf_map_write_active() indicates a map->writecnt of 0
only then we are really guaranteed to use the map's data as known constants.
For map->frozen being set and pending writes in process of still being completed
we fall back to marking that register as unknown scalar so we don't end up
making assumptions about it. With this, both TOCTOU reproducers from i) and
ii) are fixed.
Note that the map->writecnt has been converted into a atomic64 in the fix in
order to avoid a double freeze_mutex mutex_{un,}lock() pair when updating
map->writecnt in the various map update/delete BPF_* cmd flavors. Spanning
the freeze_mutex over entire map update/delete operations in syscall side
would not be possible due to then causing everything to be serialized.
Similarly, something like synchronize_rcu() after setting map->frozen to wait
for update/deletes to complete is not possible either since it would also
have to span the user copy which can sleep. On the libbpf side, this won't
break d66562fba1ce ("libbpf: Add BPF object skeleton support") as the
anonymous mmap()-ed "map initialization image" is remapped as a BPF map-backed
mmap()-ed memory where for .rodata it's non-writable.
Fixes: a23740ec43ba ("bpf: Track contents of read-only maps as scalars")
Reported-by: w1tcher.bupt(a)gmail.com
Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net>
Acked-by: Andrii Nakryiko <andrii(a)kernel.org>
Signed-off-by: Alexei Starovoitov <ast(a)kernel.org>
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index f715e8863f4d..e7a163a3146b 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -193,7 +193,7 @@ struct bpf_map {
atomic64_t usercnt;
struct work_struct work;
struct mutex freeze_mutex;
- u64 writecnt; /* writable mmap cnt; protected by freeze_mutex */
+ atomic64_t writecnt;
};
static inline bool map_value_has_spin_lock(const struct bpf_map *map)
@@ -1419,6 +1419,7 @@ void bpf_map_put(struct bpf_map *map);
void *bpf_map_area_alloc(u64 size, int numa_node);
void *bpf_map_area_mmapable_alloc(u64 size, int numa_node);
void bpf_map_area_free(void *base);
+bool bpf_map_write_active(const struct bpf_map *map);
void bpf_map_init_from_attr(struct bpf_map *map, union bpf_attr *attr);
int generic_map_lookup_batch(struct bpf_map *map,
const union bpf_attr *attr,
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 50f96ea4452a..1033ee8c0caf 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -132,6 +132,21 @@ static struct bpf_map *find_and_alloc_map(union bpf_attr *attr)
return map;
}
+static void bpf_map_write_active_inc(struct bpf_map *map)
+{
+ atomic64_inc(&map->writecnt);
+}
+
+static void bpf_map_write_active_dec(struct bpf_map *map)
+{
+ atomic64_dec(&map->writecnt);
+}
+
+bool bpf_map_write_active(const struct bpf_map *map)
+{
+ return atomic64_read(&map->writecnt) != 0;
+}
+
static u32 bpf_map_value_size(const struct bpf_map *map)
{
if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
@@ -601,11 +616,8 @@ static void bpf_map_mmap_open(struct vm_area_struct *vma)
{
struct bpf_map *map = vma->vm_file->private_data;
- if (vma->vm_flags & VM_MAYWRITE) {
- mutex_lock(&map->freeze_mutex);
- map->writecnt++;
- mutex_unlock(&map->freeze_mutex);
- }
+ if (vma->vm_flags & VM_MAYWRITE)
+ bpf_map_write_active_inc(map);
}
/* called for all unmapped memory region (including initial) */
@@ -613,11 +625,8 @@ static void bpf_map_mmap_close(struct vm_area_struct *vma)
{
struct bpf_map *map = vma->vm_file->private_data;
- if (vma->vm_flags & VM_MAYWRITE) {
- mutex_lock(&map->freeze_mutex);
- map->writecnt--;
- mutex_unlock(&map->freeze_mutex);
- }
+ if (vma->vm_flags & VM_MAYWRITE)
+ bpf_map_write_active_dec(map);
}
static const struct vm_operations_struct bpf_map_default_vmops = {
@@ -668,7 +677,7 @@ static int bpf_map_mmap(struct file *filp, struct vm_area_struct *vma)
goto out;
if (vma->vm_flags & VM_MAYWRITE)
- map->writecnt++;
+ bpf_map_write_active_inc(map);
out:
mutex_unlock(&map->freeze_mutex);
return err;
@@ -1139,6 +1148,7 @@ static int map_update_elem(union bpf_attr *attr, bpfptr_t uattr)
map = __bpf_map_get(f);
if (IS_ERR(map))
return PTR_ERR(map);
+ bpf_map_write_active_inc(map);
if (!(map_get_sys_perms(map, f) & FMODE_CAN_WRITE)) {
err = -EPERM;
goto err_put;
@@ -1174,6 +1184,7 @@ static int map_update_elem(union bpf_attr *attr, bpfptr_t uattr)
free_key:
kvfree(key);
err_put:
+ bpf_map_write_active_dec(map);
fdput(f);
return err;
}
@@ -1196,6 +1207,7 @@ static int map_delete_elem(union bpf_attr *attr)
map = __bpf_map_get(f);
if (IS_ERR(map))
return PTR_ERR(map);
+ bpf_map_write_active_inc(map);
if (!(map_get_sys_perms(map, f) & FMODE_CAN_WRITE)) {
err = -EPERM;
goto err_put;
@@ -1226,6 +1238,7 @@ static int map_delete_elem(union bpf_attr *attr)
out:
kvfree(key);
err_put:
+ bpf_map_write_active_dec(map);
fdput(f);
return err;
}
@@ -1533,6 +1546,7 @@ static int map_lookup_and_delete_elem(union bpf_attr *attr)
map = __bpf_map_get(f);
if (IS_ERR(map))
return PTR_ERR(map);
+ bpf_map_write_active_inc(map);
if (!(map_get_sys_perms(map, f) & FMODE_CAN_READ) ||
!(map_get_sys_perms(map, f) & FMODE_CAN_WRITE)) {
err = -EPERM;
@@ -1597,6 +1611,7 @@ static int map_lookup_and_delete_elem(union bpf_attr *attr)
free_key:
kvfree(key);
err_put:
+ bpf_map_write_active_dec(map);
fdput(f);
return err;
}
@@ -1624,8 +1639,7 @@ static int map_freeze(const union bpf_attr *attr)
}
mutex_lock(&map->freeze_mutex);
-
- if (map->writecnt) {
+ if (bpf_map_write_active(map)) {
err = -EBUSY;
goto err_put;
}
@@ -4171,6 +4185,9 @@ static int bpf_map_do_batch(const union bpf_attr *attr,
union bpf_attr __user *uattr,
int cmd)
{
+ bool has_read = cmd == BPF_MAP_LOOKUP_BATCH ||
+ cmd == BPF_MAP_LOOKUP_AND_DELETE_BATCH;
+ bool has_write = cmd != BPF_MAP_LOOKUP_BATCH;
struct bpf_map *map;
int err, ufd;
struct fd f;
@@ -4183,16 +4200,13 @@ static int bpf_map_do_batch(const union bpf_attr *attr,
map = __bpf_map_get(f);
if (IS_ERR(map))
return PTR_ERR(map);
-
- if ((cmd == BPF_MAP_LOOKUP_BATCH ||
- cmd == BPF_MAP_LOOKUP_AND_DELETE_BATCH) &&
- !(map_get_sys_perms(map, f) & FMODE_CAN_READ)) {
+ if (has_write)
+ bpf_map_write_active_inc(map);
+ if (has_read && !(map_get_sys_perms(map, f) & FMODE_CAN_READ)) {
err = -EPERM;
goto err_put;
}
-
- if (cmd != BPF_MAP_LOOKUP_BATCH &&
- !(map_get_sys_perms(map, f) & FMODE_CAN_WRITE)) {
+ if (has_write && !(map_get_sys_perms(map, f) & FMODE_CAN_WRITE)) {
err = -EPERM;
goto err_put;
}
@@ -4205,8 +4219,9 @@ static int bpf_map_do_batch(const union bpf_attr *attr,
BPF_DO_BATCH(map->ops->map_update_batch);
else
BPF_DO_BATCH(map->ops->map_delete_batch);
-
err_put:
+ if (has_write)
+ bpf_map_write_active_dec(map);
fdput(f);
return err;
}
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 65d2f93b7030..50efda51515b 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -4056,7 +4056,22 @@ static void coerce_reg_to_size(struct bpf_reg_state *reg, int size)
static bool bpf_map_is_rdonly(const struct bpf_map *map)
{
- return (map->map_flags & BPF_F_RDONLY_PROG) && map->frozen;
+ /* A map is considered read-only if the following condition are true:
+ *
+ * 1) BPF program side cannot change any of the map content. The
+ * BPF_F_RDONLY_PROG flag is throughout the lifetime of a map
+ * and was set at map creation time.
+ * 2) The map value(s) have been initialized from user space by a
+ * loader and then "frozen", such that no new map update/delete
+ * operations from syscall side are possible for the rest of
+ * the map's lifetime from that point onwards.
+ * 3) Any parallel/pending map update/delete operations from syscall
+ * side have been completed. Only after that point, it's safe to
+ * assume that map value(s) are immutable.
+ */
+ return (map->map_flags & BPF_F_RDONLY_PROG) &&
+ READ_ONCE(map->frozen) &&
+ !bpf_map_write_active(map);
}
static int bpf_map_direct_read(struct bpf_map *map, int off, int size, u64 *val)