mirror/freebsd
1
0
Fork 0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-17 10:26:15 +00:00
Code Issues Releases Activity

ktrace_resize_pool() locking slightly reworked:

Browse Source 
1) do not take a lock around the single atomic operation.
2) do not lose the invariant of lock by dropping/acquiring
   ktrace_mtx around free() or malloc().

MFC after:	1 Month.
This commit is contained in:
Dmitry Chagin 2011-02-25 22:03:28 +00:00
parent 2960733f48
commit b4c20e5e37
Notes: svn2git 2020-12-20 02:59:44 +00:00 
svn path=/head/; revision=219041
1 changed files with 16 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
sys/kern/kern_ktrace.c
View File

@ -133,7 +133,7 @@ static struct sx ktrace_sx;
static void ktrace_init(void *dummy);
static int sysctl_kern_ktrace_request_pool(SYSCTL_HANDLER_ARGS);
static u_int ktrace_resize_pool(u_int newsize);
static u_int ktrace_resize_pool(u_int oldsize, u_int newsize);
static struct ktr_request *ktr_getrequest(int type);
static void ktr_submitrequest(struct thread *td, struct ktr_request *req);
static void ktr_freeproc(struct proc *p, struct ucred **uc,
@ -199,9 +199,7 @@ sysctl_kern_ktrace_request_pool(SYSCTL_HANDLER_ARGS)
/* Handle easy read-only case first to avoid warnings from GCC. */
if (!req->newptr) {
mtx_lock(&ktrace_mtx);
oldsize = ktr_requestpool;
mtx_unlock(&ktrace_mtx);
return (SYSCTL_OUT(req, &oldsize, sizeof(u_int)));
}
@ -210,10 +208,8 @@ sysctl_kern_ktrace_request_pool(SYSCTL_HANDLER_ARGS)
return (error);
td = curthread;
ktrace_enter(td);
mtx_lock(&ktrace_mtx);
oldsize = ktr_requestpool;
newsize = ktrace_resize_pool(wantsize);
mtx_unlock(&ktrace_mtx);
newsize = ktrace_resize_pool(oldsize, wantsize);
ktrace_exit(td);
error = SYSCTL_OUT(req, &oldsize, sizeof(u_int));
if (error)
@ -227,38 +223,40 @@ SYSCTL_PROC(_kern_ktrace, OID_AUTO, request_pool, CTLTYPE_UINT|CTLFLAG_RW,
"Pool buffer size for ktrace(1)");
static u_int
ktrace_resize_pool(u_int newsize)
ktrace_resize_pool(u_int oldsize, u_int newsize)
{
STAILQ_HEAD(, ktr_request) ktr_new;
struct ktr_request *req;
int bound;
mtx_assert(&ktrace_mtx, MA_OWNED);
print_message = 1;
bound = newsize - ktr_requestpool;
bound = newsize - oldsize;
if (bound == 0)
return (ktr_requestpool);
if (bound < 0)
if (bound < 0) {
mtx_lock(&ktrace_mtx);
/* Shrink pool down to newsize if possible. */
while (bound++ < 0) {
req = STAILQ_FIRST(&ktr_free);
if (req == NULL)
return (ktr_requestpool);
break;
STAILQ_REMOVE_HEAD(&ktr_free, ktr_list);
ktr_requestpool--;
mtx_unlock(&ktrace_mtx);
free(req, M_KTRACE);
mtx_lock(&ktrace_mtx);
}
else
} else {
/* Grow pool up to newsize. */
STAILQ_INIT(&ktr_new);
while (bound-- > 0) {
mtx_unlock(&ktrace_mtx);
req = malloc(sizeof(struct ktr_request), M_KTRACE,
M_WAITOK);
mtx_lock(&ktrace_mtx);
STAILQ_INSERT_HEAD(&ktr_free, req, ktr_list);
ktr_requestpool++;
STAILQ_INSERT_HEAD(&ktr_new, req, ktr_list);
}
mtx_lock(&ktrace_mtx);
STAILQ_CONCAT(&ktr_free, &ktr_new);
ktr_requestpool += (newsize - oldsize);
}
mtx_unlock(&ktrace_mtx);
return (ktr_requestpool);
}