* [PATCH for-next] IB/hfi1: Convert kzalloc_node and kcalloc to use kcalloc_node
@ 2018-02-01 20:37 Dennis Dalessandro
[not found] ` <20180201203727.13590.84553.stgit-9QXIwq+3FY+1XWohqUldA0EOCMrvLtNR@public.gmane.org>
0 siblings, 1 reply; 2+ messages in thread
From: Dennis Dalessandro @ 2018-02-01 20:37 UTC (permalink / raw)
To: jgg-uk2M96/98Pc, dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA, Mike Marciniszyn,
Kamenee Arumugam
From: Kamenee Arumugam <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Kzalloc_node API doesn't check for overflows in size multiplication.
While kcalloc API check for overflows in size multiplication
but these implementations are not NUMA-aware.
This conversion allowed for correcting an allocation used in the hot
path to be on the local NUMA and ensure us overflow free multiplication
for the size of a memory allocation.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Kamenee Arumugam <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/hfi1/init.c | 17 +++++++++--------
drivers/infiniband/hw/hfi1/pio.c | 15 ++++++++-------
drivers/infiniband/hw/hfi1/sdma.c | 3 ++-
3 files changed, 19 insertions(+), 16 deletions(-)
diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index 8c4f040..33eba23 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -172,7 +172,7 @@ int hfi1_create_kctxts(struct hfi1_devdata *dd)
u16 i;
int ret;
- dd->rcd = kzalloc_node(dd->num_rcv_contexts * sizeof(*dd->rcd),
+ dd->rcd = kcalloc_node(dd->num_rcv_contexts, sizeof(*dd->rcd),
GFP_KERNEL, dd->node);
if (!dd->rcd)
return -ENOMEM;
@@ -439,15 +439,16 @@ int hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, int numa,
* The resulting value will be rounded down to the closest
* multiple of dd->rcv_entries.group_size.
*/
- rcd->egrbufs.buffers = kzalloc_node(
- rcd->egrbufs.count * sizeof(*rcd->egrbufs.buffers),
- GFP_KERNEL, numa);
+ rcd->egrbufs.buffers =
+ kcalloc_node(rcd->egrbufs.count,
+ sizeof(*rcd->egrbufs.buffers),
+ GFP_KERNEL, numa);
if (!rcd->egrbufs.buffers)
goto bail;
- rcd->egrbufs.rcvtids = kzalloc_node(
- rcd->egrbufs.count *
- sizeof(*rcd->egrbufs.rcvtids),
- GFP_KERNEL, numa);
+ rcd->egrbufs.rcvtids =
+ kcalloc_node(rcd->egrbufs.count,
+ sizeof(*rcd->egrbufs.rcvtids),
+ GFP_KERNEL, numa);
if (!rcd->egrbufs.rcvtids)
goto bail;
rcd->egrbufs.size = eager_buffer_size;
diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 4c1198b..40dac4d 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -455,8 +455,8 @@ int init_send_contexts(struct hfi1_devdata *dd)
dd->hw_to_sw = kmalloc_array(TXE_NUM_CONTEXTS, sizeof(u8),
GFP_KERNEL);
dd->send_contexts = kcalloc(dd->num_send_contexts,
- sizeof(struct send_context_info),
- GFP_KERNEL);
+ sizeof(struct send_context_info),
+ GFP_KERNEL);
if (!dd->send_contexts || !dd->hw_to_sw) {
kfree(dd->hw_to_sw);
kfree(dd->send_contexts);
@@ -856,8 +856,9 @@ struct send_context *sc_alloc(struct hfi1_devdata *dd, int type,
* so head == tail can mean empty.
*/
sc->sr_size = sci->credits + 1;
- sc->sr = kzalloc_node(sizeof(union pio_shadow_ring) *
- sc->sr_size, GFP_KERNEL, numa);
+ sc->sr = kcalloc_node(sc->sr_size,
+ sizeof(union pio_shadow_ring),
+ GFP_KERNEL, numa);
if (!sc->sr) {
sc_free(sc);
return NULL;
@@ -1958,9 +1959,9 @@ int init_pervl_scs(struct hfi1_devdata *dd)
hfi1_init_ctxt(dd->vld[15].sc);
dd->vld[15].mtu = enum_to_mtu(OPA_MTU_2048);
- dd->kernel_send_context = kzalloc_node(dd->num_send_contexts *
- sizeof(struct send_context *),
- GFP_KERNEL, dd->node);
+ dd->kernel_send_context = kcalloc_node(dd->num_send_contexts,
+ sizeof(struct send_context *),
+ GFP_KERNEL, dd->node);
if (!dd->kernel_send_context)
goto freesc15;
diff --git a/drivers/infiniband/hw/hfi1/sdma.c b/drivers/infiniband/hw/hfi1/sdma.c
index 37424a8..d8ddbfd 100644
--- a/drivers/infiniband/hw/hfi1/sdma.c
+++ b/drivers/infiniband/hw/hfi1/sdma.c
@@ -1389,7 +1389,8 @@ int sdma_init(struct hfi1_devdata *dd, u8 port)
num_engines, descq_cnt);
/* alloc memory for array of send engines */
- dd->per_sdma = kcalloc(num_engines, sizeof(*dd->per_sdma), GFP_KERNEL);
+ dd->per_sdma = kcalloc_node(num_engines, sizeof(*dd->per_sdma),
+ GFP_KERNEL, dd->node);
if (!dd->per_sdma)
return ret;
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH for-next] IB/hfi1: Convert kzalloc_node and kcalloc to use kcalloc_node
[not found] ` <20180201203727.13590.84553.stgit-9QXIwq+3FY+1XWohqUldA0EOCMrvLtNR@public.gmane.org>
@ 2018-02-01 22:52 ` Jason Gunthorpe
0 siblings, 0 replies; 2+ messages in thread
From: Jason Gunthorpe @ 2018-02-01 22:52 UTC (permalink / raw)
To: Dennis Dalessandro
Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA, Mike Marciniszyn,
Kamenee Arumugam
On Thu, Feb 01, 2018 at 12:37:30PM -0800, Dennis Dalessandro wrote:
> From: Kamenee Arumugam <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>
> Kzalloc_node API doesn't check for overflows in size multiplication.
> While kcalloc API check for overflows in size multiplication
> but these implementations are not NUMA-aware.
>
> This conversion allowed for correcting an allocation used in the hot
> path to be on the local NUMA and ensure us overflow free multiplication
> for the size of a memory allocation.
>
> Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Kamenee Arumugam <kamenee.arumugam-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> drivers/infiniband/hw/hfi1/init.c | 17 +++++++++--------
> drivers/infiniband/hw/hfi1/pio.c | 15 ++++++++-------
> drivers/infiniband/hw/hfi1/sdma.c | 3 ++-
> 3 files changed, 19 insertions(+), 16 deletions(-)
Applied to for-next
Thanks,
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2018-02-01 22:52 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-02-01 20:37 [PATCH for-next] IB/hfi1: Convert kzalloc_node and kcalloc to use kcalloc_node Dennis Dalessandro
[not found] ` <20180201203727.13590.84553.stgit-9QXIwq+3FY+1XWohqUldA0EOCMrvLtNR@public.gmane.org>
2018-02-01 22:52 ` Jason Gunthorpe
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox