* [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing
@ 2019-01-14 14:10 Yuval Shaia
2019-01-14 15:11 ` Philippe Mathieu-Daudé
0 siblings, 1 reply; 4+ messages in thread
From: Yuval Shaia @ 2019-01-14 14:10 UTC (permalink / raw)
To: yuval.shaia, marcel.apfelbaum, qemu-devel
Make sure objects are not NULL before calling to non-nulll-safe
destructors.
Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
---
hw/rdma/rdma_backend.c | 6 ++++--
hw/rdma/rdma_rm.c | 7 ++++++-
2 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/hw/rdma/rdma_backend.c b/hw/rdma/rdma_backend.c
index b49edaacaf..3ee5172c96 100644
--- a/hw/rdma/rdma_backend.c
+++ b/hw/rdma/rdma_backend.c
@@ -1066,8 +1066,10 @@ static void mad_fini(RdmaBackendDev *backend_dev)
pr_dbg("Stopping MAD\n");
disable_rdmacm_mux_async(backend_dev);
qemu_chr_fe_disconnect(backend_dev->rdmacm_mux.chr_be);
- qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
- qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
+ if (backend_dev->recv_mads_list.list) {
+ qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
+ qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
+ }
}
int rdma_backend_get_gid_index(RdmaBackendDev *backend_dev,
diff --git a/hw/rdma/rdma_rm.c b/hw/rdma/rdma_rm.c
index f5b1295890..8bf241e91f 100644
--- a/hw/rdma/rdma_rm.c
+++ b/hw/rdma/rdma_rm.c
@@ -41,6 +41,9 @@ static inline void res_tbl_init(const char *name, RdmaRmResTbl *tbl,
static inline void res_tbl_free(RdmaRmResTbl *tbl)
{
+ if (!tbl->bitmap) {
+ return;
+ }
qemu_mutex_destroy(&tbl->lock);
g_free(tbl->tbl);
g_free(tbl->bitmap);
@@ -655,5 +658,7 @@ void rdma_rm_fini(RdmaDeviceResources *dev_res, RdmaBackendDev *backend_dev,
res_tbl_free(&dev_res->cq_tbl);
res_tbl_free(&dev_res->pd_tbl);
- g_hash_table_destroy(dev_res->qp_hash);
+ if (dev_res->qp_hash) {
+ g_hash_table_destroy(dev_res->qp_hash);
+ }
}
--
2.17.2
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing
2019-01-14 14:10 [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing Yuval Shaia
@ 2019-01-14 15:11 ` Philippe Mathieu-Daudé
2019-01-16 10:03 ` Yuval Shaia
0 siblings, 1 reply; 4+ messages in thread
From: Philippe Mathieu-Daudé @ 2019-01-14 15:11 UTC (permalink / raw)
To: Yuval Shaia, marcel.apfelbaum, qemu-devel
Hi Yuval,
On 1/14/19 3:10 PM, Yuval Shaia wrote:
> Make sure objects are not NULL before calling to non-nulll-safe
null
> destructors.
>
> Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
> ---
> hw/rdma/rdma_backend.c | 6 ++++--
> hw/rdma/rdma_rm.c | 7 ++++++-
> 2 files changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/hw/rdma/rdma_backend.c b/hw/rdma/rdma_backend.c
> index b49edaacaf..3ee5172c96 100644
> --- a/hw/rdma/rdma_backend.c
> +++ b/hw/rdma/rdma_backend.c
> @@ -1066,8 +1066,10 @@ static void mad_fini(RdmaBackendDev *backend_dev)
> pr_dbg("Stopping MAD\n");
> disable_rdmacm_mux_async(backend_dev);
> qemu_chr_fe_disconnect(backend_dev->rdmacm_mux.chr_be);
> - qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
> - qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
> + if (backend_dev->recv_mads_list.list) {
This is only reachable when the backend didn't success at initializing,
right?
> + qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
> + qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
> + }
> }
>
> int rdma_backend_get_gid_index(RdmaBackendDev *backend_dev,
> diff --git a/hw/rdma/rdma_rm.c b/hw/rdma/rdma_rm.c
> index f5b1295890..8bf241e91f 100644
> --- a/hw/rdma/rdma_rm.c
> +++ b/hw/rdma/rdma_rm.c
> @@ -41,6 +41,9 @@ static inline void res_tbl_init(const char *name, RdmaRmResTbl *tbl,
>
> static inline void res_tbl_free(RdmaRmResTbl *tbl)
> {
> + if (!tbl->bitmap) {
> + return;
> + }
> qemu_mutex_destroy(&tbl->lock);
> g_free(tbl->tbl);
> g_free(tbl->bitmap);
> @@ -655,5 +658,7 @@ void rdma_rm_fini(RdmaDeviceResources *dev_res, RdmaBackendDev *backend_dev,
> res_tbl_free(&dev_res->cq_tbl);
> res_tbl_free(&dev_res->pd_tbl);
>
> - g_hash_table_destroy(dev_res->qp_hash);
> + if (dev_res->qp_hash) {
> + g_hash_table_destroy(dev_res->qp_hash);
> + }
> }
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing
2019-01-14 15:11 ` Philippe Mathieu-Daudé
@ 2019-01-16 10:03 ` Yuval Shaia
2019-01-16 14:48 ` Philippe Mathieu-Daudé
0 siblings, 1 reply; 4+ messages in thread
From: Yuval Shaia @ 2019-01-16 10:03 UTC (permalink / raw)
To: Philippe Mathieu-Daudé; +Cc: marcel.apfelbaum, qemu-devel
On Mon, Jan 14, 2019 at 04:11:22PM +0100, Philippe Mathieu-Daudé wrote:
> Hi Yuval,
>
> On 1/14/19 3:10 PM, Yuval Shaia wrote:
> > Make sure objects are not NULL before calling to non-nulll-safe
>
> null
will fix
>
> > destructors.
> >
> > Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
> > ---
> > hw/rdma/rdma_backend.c | 6 ++++--
> > hw/rdma/rdma_rm.c | 7 ++++++-
> > 2 files changed, 10 insertions(+), 3 deletions(-)
> >
> > diff --git a/hw/rdma/rdma_backend.c b/hw/rdma/rdma_backend.c
> > index b49edaacaf..3ee5172c96 100644
> > --- a/hw/rdma/rdma_backend.c
> > +++ b/hw/rdma/rdma_backend.c
> > @@ -1066,8 +1066,10 @@ static void mad_fini(RdmaBackendDev *backend_dev)
> > pr_dbg("Stopping MAD\n");
> > disable_rdmacm_mux_async(backend_dev);
> > qemu_chr_fe_disconnect(backend_dev->rdmacm_mux.chr_be);
> > - qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
> > - qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
> > + if (backend_dev->recv_mads_list.list) {
>
> This is only reachable when the backend didn't success at initializing,
> right?
Yes, just to cover the case where fini() was called even when init() fails.
>
> > + qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
> > + qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
> > + }
> > }
> >
> > int rdma_backend_get_gid_index(RdmaBackendDev *backend_dev,
> > diff --git a/hw/rdma/rdma_rm.c b/hw/rdma/rdma_rm.c
> > index f5b1295890..8bf241e91f 100644
> > --- a/hw/rdma/rdma_rm.c
> > +++ b/hw/rdma/rdma_rm.c
> > @@ -41,6 +41,9 @@ static inline void res_tbl_init(const char *name, RdmaRmResTbl *tbl,
> >
> > static inline void res_tbl_free(RdmaRmResTbl *tbl)
> > {
> > + if (!tbl->bitmap) {
> > + return;
> > + }
> > qemu_mutex_destroy(&tbl->lock);
> > g_free(tbl->tbl);
> > g_free(tbl->bitmap);
> > @@ -655,5 +658,7 @@ void rdma_rm_fini(RdmaDeviceResources *dev_res, RdmaBackendDev *backend_dev,
> > res_tbl_free(&dev_res->cq_tbl);
> > res_tbl_free(&dev_res->pd_tbl);
> >
> > - g_hash_table_destroy(dev_res->qp_hash);
> > + if (dev_res->qp_hash) {
> > + g_hash_table_destroy(dev_res->qp_hash);
> > + }
> > }
> >
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing
2019-01-16 10:03 ` Yuval Shaia
@ 2019-01-16 14:48 ` Philippe Mathieu-Daudé
0 siblings, 0 replies; 4+ messages in thread
From: Philippe Mathieu-Daudé @ 2019-01-16 14:48 UTC (permalink / raw)
To: Yuval Shaia; +Cc: marcel.apfelbaum, qemu-devel
On 1/16/19 11:03 AM, Yuval Shaia wrote:
> On Mon, Jan 14, 2019 at 04:11:22PM +0100, Philippe Mathieu-Daudé wrote:
>> Hi Yuval,
>>
>> On 1/14/19 3:10 PM, Yuval Shaia wrote:
>>> Make sure objects are not NULL before calling to non-nulll-safe
>>
>> null
>
> will fix
>
>>
>>> destructors.
>>>
>>> Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
>>> ---
>>> hw/rdma/rdma_backend.c | 6 ++++--
>>> hw/rdma/rdma_rm.c | 7 ++++++-
>>> 2 files changed, 10 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/hw/rdma/rdma_backend.c b/hw/rdma/rdma_backend.c
>>> index b49edaacaf..3ee5172c96 100644
>>> --- a/hw/rdma/rdma_backend.c
>>> +++ b/hw/rdma/rdma_backend.c
>>> @@ -1066,8 +1066,10 @@ static void mad_fini(RdmaBackendDev *backend_dev)
>>> pr_dbg("Stopping MAD\n");
>>> disable_rdmacm_mux_async(backend_dev);
>>> qemu_chr_fe_disconnect(backend_dev->rdmacm_mux.chr_be);
>>> - qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
>>> - qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
>>> + if (backend_dev->recv_mads_list.list) {
>>
>> This is only reachable when the backend didn't success at initializing,
>> right?
>
> Yes, just to cover the case where fini() was called even when init() fails.
OK. Can you add a line about this in the commit description please?
With the comment updated:
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>
>>
>>> + qlist_destroy_obj(QOBJECT(backend_dev->recv_mads_list.list));
>>> + qemu_mutex_destroy(&backend_dev->recv_mads_list.lock);
>>> + }
>>> }
>>>
>>> int rdma_backend_get_gid_index(RdmaBackendDev *backend_dev,
>>> diff --git a/hw/rdma/rdma_rm.c b/hw/rdma/rdma_rm.c
>>> index f5b1295890..8bf241e91f 100644
>>> --- a/hw/rdma/rdma_rm.c
>>> +++ b/hw/rdma/rdma_rm.c
>>> @@ -41,6 +41,9 @@ static inline void res_tbl_init(const char *name, RdmaRmResTbl *tbl,
>>>
>>> static inline void res_tbl_free(RdmaRmResTbl *tbl)
>>> {
>>> + if (!tbl->bitmap) {
>>> + return;
>>> + }
>>> qemu_mutex_destroy(&tbl->lock);
>>> g_free(tbl->tbl);
>>> g_free(tbl->bitmap);
>>> @@ -655,5 +658,7 @@ void rdma_rm_fini(RdmaDeviceResources *dev_res, RdmaBackendDev *backend_dev,
>>> res_tbl_free(&dev_res->cq_tbl);
>>> res_tbl_free(&dev_res->pd_tbl);
>>>
>>> - g_hash_table_destroy(dev_res->qp_hash);
>>> + if (dev_res->qp_hash) {
>>> + g_hash_table_destroy(dev_res->qp_hash);
>>> + }
>>> }
>>>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2019-01-16 14:48 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-01-14 14:10 [Qemu-devel] [PATCH] hw/rdma: Verify that ptr is not NULL before freeing Yuval Shaia
2019-01-14 15:11 ` Philippe Mathieu-Daudé
2019-01-16 10:03 ` Yuval Shaia
2019-01-16 14:48 ` Philippe Mathieu-Daudé
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.