diff mbox series

hw: pvrdma: fix memory leak in error path

Message ID 20190103130338.49081-1-liq3ea@163.com (mailing list archive)
State New, archived
Headers show
Series hw: pvrdma: fix memory leak in error path | expand

Commit Message

Li Qiang Jan. 3, 2019, 1:03 p.m. UTC 
Spotted by Coverity: CID 1398595

Signed-off-by: Li Qiang <liq3ea@163.com>
---
 hw/rdma/vmw/pvrdma_qp_ops.c | 2 ++
 1 file changed, 2 insertions(+)

Comments

Philippe Mathieu-Daudé Jan. 3, 2019, 1:47 p.m. UTC | #1 
On 1/3/19 2:03 PM, Li Qiang wrote:
> Spotted by Coverity: CID 1398595
> 

Fixes: 2b05705dc8

> Signed-off-by: Li Qiang <liq3ea@163.com>

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

> ---
>  hw/rdma/vmw/pvrdma_qp_ops.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/hw/rdma/vmw/pvrdma_qp_ops.c b/hw/rdma/vmw/pvrdma_qp_ops.c
> index 300471a4c9..584be2043e 100644
> --- a/hw/rdma/vmw/pvrdma_qp_ops.c
> +++ b/hw/rdma/vmw/pvrdma_qp_ops.c
> @@ -168,6 +168,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>          sgid = rdma_rm_get_gid(&dev->rdma_dev_res, wqe->hdr.wr.ud.av.gid_index);
>          if (!sgid) {
>              pr_dbg("Fail to get gid for idx %d\n", wqe->hdr.wr.ud.av.gid_index);
> +            g_free(comp_ctx);
>              return -EIO;
>          }
>          pr_dbg("sgid_id=%d, sgid=0x%llx\n", wqe->hdr.wr.ud.av.gid_index,
> @@ -179,6 +180,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>          if (sgid_idx <= 0) {
>              pr_dbg("Fail to get bk sgid_idx for sgid_idx %d\n",
>                     wqe->hdr.wr.ud.av.gid_index);
> +            g_free(comp_ctx);
>              return -EIO;
>          }
>  
>
Yuval Shaia Jan. 7, 2019, 4:26 p.m. UTC | #2 
On Thu, Jan 03, 2019 at 02:47:37PM +0100, Philippe Mathieu-Daudé wrote:
> On 1/3/19 2:03 PM, Li Qiang wrote:
> > Spotted by Coverity: CID 1398595
> > 
> 
> Fixes: 2b05705dc8
> 
> > Signed-off-by: Li Qiang <liq3ea@163.com>
> 
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> 
> > ---
> >  hw/rdma/vmw/pvrdma_qp_ops.c | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/hw/rdma/vmw/pvrdma_qp_ops.c b/hw/rdma/vmw/pvrdma_qp_ops.c
> > index 300471a4c9..584be2043e 100644
> > --- a/hw/rdma/vmw/pvrdma_qp_ops.c
> > +++ b/hw/rdma/vmw/pvrdma_qp_ops.c
> > @@ -168,6 +168,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
> >          sgid = rdma_rm_get_gid(&dev->rdma_dev_res, wqe->hdr.wr.ud.av.gid_index);
> >          if (!sgid) {
> >              pr_dbg("Fail to get gid for idx %d\n", wqe->hdr.wr.ud.av.gid_index);
> > +            g_free(comp_ctx);
> >              return -EIO;
> >          }
> >          pr_dbg("sgid_id=%d, sgid=0x%llx\n", wqe->hdr.wr.ud.av.gid_index,
> > @@ -179,6 +180,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
> >          if (sgid_idx <= 0) {
> >              pr_dbg("Fail to get bk sgid_idx for sgid_idx %d\n",
> >                     wqe->hdr.wr.ud.av.gid_index);
> > +            g_free(comp_ctx);
> >              return -EIO;
> >          }

Since comp_ctx is not used until the two checks are done we just can
relocate the allocation & initialization right after the two checks.

Yuval

> >  
> >
Li Qiang Jan. 8, 2019, 6:40 a.m. UTC | #3 
At 2019-01-08 00:26:58, "Yuval Shaia" <yuval.shaia@oracle.com> wrote:
>On Thu, Jan 03, 2019 at 02:47:37PM +0100, Philippe Mathieu-Daudé wrote:
>> On 1/3/19 2:03 PM, Li Qiang wrote:
>> > Spotted by Coverity: CID 1398595
>> > 
>> 
>> Fixes: 2b05705dc8
>> 
>> > Signed-off-by: Li Qiang <liq3ea@163.com>
>> 
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> 
>> > ---
>> >  hw/rdma/vmw/pvrdma_qp_ops.c | 2 ++
>> >  1 file changed, 2 insertions(+)
>> > 
>> > diff --git a/hw/rdma/vmw/pvrdma_qp_ops.c b/hw/rdma/vmw/pvrdma_qp_ops.c
>> > index 300471a4c9..584be2043e 100644
>> > --- a/hw/rdma/vmw/pvrdma_qp_ops.c
>> > +++ b/hw/rdma/vmw/pvrdma_qp_ops.c
>> > @@ -168,6 +168,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>> >          sgid = rdma_rm_get_gid(&dev->rdma_dev_res, wqe->hdr.wr.ud.av.gid_index);
>> >          if (!sgid) {
>> >              pr_dbg("Fail to get gid for idx %d\n", wqe->hdr.wr.ud.av.gid_index);
>> > +            g_free(comp_ctx);
>> >              return -EIO;
>> >          }
>> >          pr_dbg("sgid_id=%d, sgid=0x%llx\n", wqe->hdr.wr.ud.av.gid_index,
>> > @@ -179,6 +180,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>> >          if (sgid_idx <= 0) {
>> >              pr_dbg("Fail to get bk sgid_idx for sgid_idx %d\n",
>> >                     wqe->hdr.wr.ud.av.gid_index);
>> > +            g_free(comp_ctx);
>> >              return -EIO;
>> >          }
>
>Since comp_ctx is not used until the two checks are done we just can
>relocate the allocation & initialization right after the two checks.

>


OK, will send a revised version later.


Thanks,
Li Qiang


>Yuval
>
>> >  
>> >
Philippe Mathieu-Daudé Jan. 8, 2019, 5:47 p.m. UTC | #4 
On 1/7/19 5:26 PM, Yuval Shaia wrote:
> On Thu, Jan 03, 2019 at 02:47:37PM +0100, Philippe Mathieu-Daudé wrote:
>> On 1/3/19 2:03 PM, Li Qiang wrote:
>>> Spotted by Coverity: CID 1398595
>>>
>>
>> Fixes: 2b05705dc8
>>
>>> Signed-off-by: Li Qiang <liq3ea@163.com>
>>
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>>
>>> ---
>>>  hw/rdma/vmw/pvrdma_qp_ops.c | 2 ++
>>>  1 file changed, 2 insertions(+)
>>>
>>> diff --git a/hw/rdma/vmw/pvrdma_qp_ops.c b/hw/rdma/vmw/pvrdma_qp_ops.c
>>> index 300471a4c9..584be2043e 100644
>>> --- a/hw/rdma/vmw/pvrdma_qp_ops.c
>>> +++ b/hw/rdma/vmw/pvrdma_qp_ops.c
>>> @@ -168,6 +168,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>>>          sgid = rdma_rm_get_gid(&dev->rdma_dev_res, wqe->hdr.wr.ud.av.gid_index);
>>>          if (!sgid) {
>>>              pr_dbg("Fail to get gid for idx %d\n", wqe->hdr.wr.ud.av.gid_index);
>>> +            g_free(comp_ctx);
>>>              return -EIO;
>>>          }
>>>          pr_dbg("sgid_id=%d, sgid=0x%llx\n", wqe->hdr.wr.ud.av.gid_index,
>>> @@ -179,6 +180,7 @@ int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
>>>          if (sgid_idx <= 0) {
>>>              pr_dbg("Fail to get bk sgid_idx for sgid_idx %d\n",
>>>                     wqe->hdr.wr.ud.av.gid_index);
>>> +            g_free(comp_ctx);
>>>              return -EIO;
>>>          }
> 
> Since comp_ctx is not used until the two checks are done we just can
> relocate the allocation & initialization right after the two checks.

Oops I missed this, thanks your suggestion is way cleaner!

Phil.
diff mbox series

Patch

diff --git a/hw/rdma/vmw/pvrdma_qp_ops.c b/hw/rdma/vmw/pvrdma_qp_ops.c
index 300471a4c9..584be2043e 100644
--- a/hw/rdma/vmw/pvrdma_qp_ops.c
+++ b/hw/rdma/vmw/pvrdma_qp_ops.c
@@ -168,6 +168,7 @@  int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
         sgid = rdma_rm_get_gid(&dev->rdma_dev_res, wqe->hdr.wr.ud.av.gid_index);
         if (!sgid) {
             pr_dbg("Fail to get gid for idx %d\n", wqe->hdr.wr.ud.av.gid_index);
+            g_free(comp_ctx);
             return -EIO;
         }
         pr_dbg("sgid_id=%d, sgid=0x%llx\n", wqe->hdr.wr.ud.av.gid_index,
@@ -179,6 +180,7 @@  int pvrdma_qp_send(PVRDMADev *dev, uint32_t qp_handle)
         if (sgid_idx <= 0) {
             pr_dbg("Fail to get bk sgid_idx for sgid_idx %d\n",
                    wqe->hdr.wr.ud.av.gid_index);
+            g_free(comp_ctx);
             return -EIO;
         }