Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux

lockd: convert nlm_host.h_count from atomic_t to refcount_t

atomic_t variables are currently used to implement reference
counters with the following properties:
- counter is initialized to 1 using atomic_set()
- a resource is freed upon counter reaching zero
- once counter reaches zero, its further
increments aren't allowed
- counter schema uses basic atomic operations
(set, inc, inc_not_zero, dec_and_test, etc.)

Such atomic variables should be converted to a newly provided
refcount_t type and API that prevents accidental counter overflows
and underflows. This is important since overflows and underflows
can lead to use-after-free situation and be exploitable.

The variable nlm_host.h_count is used as pure reference counter.
Convert it to refcount_t and fix up the operations.

**Important note for maintainers:

Some functions from refcount_t API defined in lib/refcount.c
have different memory ordering guarantees than their atomic
counterparts.
The full comparison can be seen in
https://lkml.org/lkml/2017/11/15/57 and it is hopefully soon
in state to be merged to the documentation tree.
Normally the differences should not matter since refcount_t provides
enough guarantees to satisfy the refcounting use cases, but in
some rare cases it might matter.
Please double check that you don't have some undocumented
memory guarantees for this variable usage.

For the nlm_host.h_count it might make a difference
in following places:
- nlmsvc_release_host(): decrement in refcount_dec()
provides RELEASE ordering, while original atomic_dec()
was fully unordered. Since the change is for better, it
should not matter.
- nlmclnt_release_host(): decrement in refcount_dec_and_test() only
provides RELEASE ordering and control dependency on success
vs. fully ordered atomic counterpart. It doesn't seem to
matter in this case since object freeing happens under mutex
lock anyway.

Suggested-by: Kees Cook <keescook@chromium.org>
Reviewed-by: David Windsor <dwindsor@gmail.com>
Reviewed-by: Hans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: Elena Reshetova <elena.reshetova@intel.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>

authored by

Elena Reshetova and committed by
Trond Myklebust
fee21fb5 ba4a76f7

+9 -8
+7 -7
fs/lockd/host.c
··· 151 151 host->h_state = 0; 152 152 host->h_nsmstate = 0; 153 153 host->h_pidcount = 0; 154 - atomic_set(&host->h_count, 1); 154 + refcount_set(&host->h_count, 1); 155 155 mutex_init(&host->h_mutex); 156 156 host->h_nextrebind = now + NLM_HOST_REBIND; 157 157 host->h_expires = now + NLM_HOST_EXPIRE; ··· 290 290 291 291 WARN_ON_ONCE(host->h_server); 292 292 293 - if (atomic_dec_and_test(&host->h_count)) { 293 + if (refcount_dec_and_test(&host->h_count)) { 294 294 WARN_ON_ONCE(!list_empty(&host->h_lockowners)); 295 295 WARN_ON_ONCE(!list_empty(&host->h_granted)); 296 296 WARN_ON_ONCE(!list_empty(&host->h_reclaim)); ··· 410 410 dprintk("lockd: release server host %s\n", host->h_name); 411 411 412 412 WARN_ON_ONCE(!host->h_server); 413 - atomic_dec(&host->h_count); 413 + refcount_dec(&host->h_count); 414 414 } 415 415 416 416 /* ··· 504 504 { 505 505 if (host) { 506 506 dprintk("lockd: get host %s\n", host->h_name); 507 - atomic_inc(&host->h_count); 507 + refcount_inc(&host->h_count); 508 508 host->h_expires = jiffies + NLM_HOST_EXPIRE; 509 509 } 510 510 return host; ··· 593 593 if (net && host->net != net) 594 594 continue; 595 595 dprintk(" %s (cnt %d use %d exp %ld net %x)\n", 596 - host->h_name, atomic_read(&host->h_count), 596 + host->h_name, refcount_read(&host->h_count), 597 597 host->h_inuse, host->h_expires, host->net->ns.inum); 598 598 } 599 599 } ··· 662 662 for_each_host_safe(host, next, chain, nlm_server_hosts) { 663 663 if (net && host->net != net) 664 664 continue; 665 - if (atomic_read(&host->h_count) || host->h_inuse 665 + if (refcount_read(&host->h_count) || host->h_inuse 666 666 || time_before(jiffies, host->h_expires)) { 667 667 dprintk("nlm_gc_hosts skipping %s " 668 668 "(cnt %d use %d exp %ld net %x)\n", 669 - host->h_name, atomic_read(&host->h_count), 669 + host->h_name, refcount_read(&host->h_count), 670 670 host->h_inuse, host->h_expires, 671 671 host->net->ns.inum); 672 672 continue;
+2 -1
include/linux/lockd/lockd.h
··· 17 17 #include <net/ipv6.h> 18 18 #include <linux/fs.h> 19 19 #include <linux/kref.h> 20 + #include <linux/refcount.h> 20 21 #include <linux/utsname.h> 21 22 #include <linux/lockd/bind.h> 22 23 #include <linux/lockd/xdr.h> ··· 59 58 u32 h_state; /* pseudo-state counter */ 60 59 u32 h_nsmstate; /* true remote NSM state */ 61 60 u32 h_pidcount; /* Pseudopids */ 62 - atomic_t h_count; /* reference count */ 61 + refcount_t h_count; /* reference count */ 63 62 struct mutex h_mutex; /* mutex for pmap binding */ 64 63 unsigned long h_nextrebind; /* next portmap call */ 65 64 unsigned long h_expires; /* eligible for GC */