* [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context
@ 2022-05-19 7:43 Ziyang Xuan
2022-05-19 7:29 ` Loic Poulain
0 siblings, 1 reply; 5+ messages in thread
From: Ziyang Xuan @ 2022-05-19 7:43 UTC (permalink / raw)
To: chandrashekar.devegowda, linuxwwan, chiranjeevi.rapolu,
haijun.liu, m.chetan.kumar, ricardo.martinez, loic.poulain,
ryazanov.s.a, johannes, davem, edumazet, kuba, pabeni, netdev
t7xx_cldma_clear_rxq() call t7xx_cldma_alloc_and_map_skb() in spin_lock
context, But __dev_alloc_skb() in t7xx_cldma_alloc_and_map_skb() uses
GFP_KERNEL, that will introduce scheduling factor in spin_lock context.
Because t7xx_cldma_clear_rxq() is called after stopping CLDMA, so we can
remove the spin_lock from t7xx_cldma_clear_rxq().
Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface")
Suggested-by: Ricardo Martinez <ricardo.martinez@linux.intel.com>
Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
---
drivers/net/wwan/t7xx/t7xx_hif_cldma.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
---
v3:
- Add Suggested-by and simplify comments
v2:
- Remove spin_lock instead of using GFP_ATOMIC
diff --git a/drivers/net/wwan/t7xx/t7xx_hif_cldma.c b/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
index 46066dcd2607..3a46a5bea24f 100644
--- a/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
+++ b/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
@@ -782,10 +782,11 @@ static int t7xx_cldma_clear_rxq(struct cldma_ctrl *md_ctrl, int qnum)
struct cldma_queue *rxq = &md_ctrl->rxq[qnum];
struct cldma_request *req;
struct cldma_gpd *gpd;
- unsigned long flags;
int ret = 0;
- spin_lock_irqsave(&rxq->ring_lock, flags);
+ /* CLDMA has been stopped. There is not any CLDMA IRQ, holding
+ * ring_lock is not needed.
+ */
t7xx_cldma_q_reset(rxq);
list_for_each_entry(req, &rxq->tr_ring->gpd_ring, entry) {
gpd = req->gpd;
@@ -808,7 +809,6 @@ static int t7xx_cldma_clear_rxq(struct cldma_ctrl *md_ctrl, int qnum)
t7xx_cldma_gpd_set_data_ptr(req->gpd, req->mapped_buff);
}
- spin_unlock_irqrestore(&rxq->ring_lock, flags);
return ret;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context
2022-05-19 7:43 [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context Ziyang Xuan
@ 2022-05-19 7:29 ` Loic Poulain
2022-05-21 0:25 ` Jakub Kicinski
0 siblings, 1 reply; 5+ messages in thread
From: Loic Poulain @ 2022-05-19 7:29 UTC (permalink / raw)
To: Ziyang Xuan
Cc: chandrashekar.devegowda, linuxwwan, chiranjeevi.rapolu,
haijun.liu, m.chetan.kumar, ricardo.martinez, ryazanov.s.a,
johannes, davem, edumazet, kuba, pabeni, netdev
On Thu, 19 May 2022 at 09:26, Ziyang Xuan <william.xuanziyang@huawei.com> wrote:
>
> t7xx_cldma_clear_rxq() call t7xx_cldma_alloc_and_map_skb() in spin_lock
> context, But __dev_alloc_skb() in t7xx_cldma_alloc_and_map_skb() uses
> GFP_KERNEL, that will introduce scheduling factor in spin_lock context.
>
> Because t7xx_cldma_clear_rxq() is called after stopping CLDMA, so we can
> remove the spin_lock from t7xx_cldma_clear_rxq().
>
> Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface")
> Suggested-by: Ricardo Martinez <ricardo.martinez@linux.intel.com>
> Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
> ---
> drivers/net/wwan/t7xx/t7xx_hif_cldma.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
> ---
> v3:
> - Add Suggested-by and simplify comments
> v2:
> - Remove spin_lock instead of using GFP_ATOMIC
>
> diff --git a/drivers/net/wwan/t7xx/t7xx_hif_cldma.c b/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
> index 46066dcd2607..3a46a5bea24f 100644
> --- a/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
> +++ b/drivers/net/wwan/t7xx/t7xx_hif_cldma.c
> @@ -782,10 +782,11 @@ static int t7xx_cldma_clear_rxq(struct cldma_ctrl *md_ctrl, int qnum)
> struct cldma_queue *rxq = &md_ctrl->rxq[qnum];
> struct cldma_request *req;
> struct cldma_gpd *gpd;
> - unsigned long flags;
> int ret = 0;
>
> - spin_lock_irqsave(&rxq->ring_lock, flags);
> + /* CLDMA has been stopped. There is not any CLDMA IRQ, holding
> + * ring_lock is not needed.
> + */
> t7xx_cldma_q_reset(rxq);
> list_for_each_entry(req, &rxq->tr_ring->gpd_ring, entry) {
> gpd = req->gpd;
> @@ -808,7 +809,6 @@ static int t7xx_cldma_clear_rxq(struct cldma_ctrl *md_ctrl, int qnum)
>
> t7xx_cldma_gpd_set_data_ptr(req->gpd, req->mapped_buff);
> }
> - spin_unlock_irqrestore(&rxq->ring_lock, flags);
>
> return ret;
> }
> --
> 2.25.1
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context
2022-05-19 7:29 ` Loic Poulain
@ 2022-05-21 0:25 ` Jakub Kicinski
2022-05-21 1:01 ` Jakub Kicinski
0 siblings, 1 reply; 5+ messages in thread
From: Jakub Kicinski @ 2022-05-21 0:25 UTC (permalink / raw)
To: Loic Poulain
Cc: Ziyang Xuan, chandrashekar.devegowda, linuxwwan,
chiranjeevi.rapolu, haijun.liu, m.chetan.kumar, ricardo.martinez,
ryazanov.s.a, johannes, davem, edumazet, pabeni, netdev
On Thu, 19 May 2022 09:29:12 +0200 Loic Poulain wrote:
> On Thu, 19 May 2022 at 09:26, Ziyang Xuan <william.xuanziyang@huawei.com> wrote:
> >
> > t7xx_cldma_clear_rxq() call t7xx_cldma_alloc_and_map_skb() in spin_lock
> > context, But __dev_alloc_skb() in t7xx_cldma_alloc_and_map_skb() uses
> > GFP_KERNEL, that will introduce scheduling factor in spin_lock context.
> >
> > Because t7xx_cldma_clear_rxq() is called after stopping CLDMA, so we can
> > remove the spin_lock from t7xx_cldma_clear_rxq().
> >
> > Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface")
> > Suggested-by: Ricardo Martinez <ricardo.martinez@linux.intel.com>
> > Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
>
> Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
Wait, you reviewed two different fixes for the same issue?
Please say something when that happens I thought both are needed :/
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context
2022-05-21 0:25 ` Jakub Kicinski
@ 2022-05-21 1:01 ` Jakub Kicinski
2022-05-21 17:26 ` Loic Poulain
0 siblings, 1 reply; 5+ messages in thread
From: Jakub Kicinski @ 2022-05-21 1:01 UTC (permalink / raw)
To: Loic Poulain
Cc: Ziyang Xuan, chandrashekar.devegowda, linuxwwan,
chiranjeevi.rapolu, haijun.liu, m.chetan.kumar, ricardo.martinez,
ryazanov.s.a, johannes, davem, edumazet, pabeni, netdev
On Fri, 20 May 2022 17:25:56 -0700 Jakub Kicinski wrote:
> On Thu, 19 May 2022 09:29:12 +0200 Loic Poulain wrote:
> > On Thu, 19 May 2022 at 09:26, Ziyang Xuan <william.xuanziyang@huawei.com> wrote:
> > >
> > > t7xx_cldma_clear_rxq() call t7xx_cldma_alloc_and_map_skb() in spin_lock
> > > context, But __dev_alloc_skb() in t7xx_cldma_alloc_and_map_skb() uses
> > > GFP_KERNEL, that will introduce scheduling factor in spin_lock context.
> > >
> > > Because t7xx_cldma_clear_rxq() is called after stopping CLDMA, so we can
> > > remove the spin_lock from t7xx_cldma_clear_rxq().
> > >
> > > Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface")
> > > Suggested-by: Ricardo Martinez <ricardo.martinez@linux.intel.com>
> > > Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
> >
> > Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
>
> Wait, you reviewed two different fixes for the same issue?
> Please say something when that happens I thought both are needed :/
FWIW I pushed out the other one before I realized (they both apply
without conflicts so I thought they fixed different issues)
If this one is preferred please respin and squash a revert of
9ee152ee3ee3 into it.
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context
2022-05-21 1:01 ` Jakub Kicinski
@ 2022-05-21 17:26 ` Loic Poulain
0 siblings, 0 replies; 5+ messages in thread
From: Loic Poulain @ 2022-05-21 17:26 UTC (permalink / raw)
To: Jakub Kicinski
Cc: Ziyang Xuan, chandrashekar.devegowda, chiranjeevi.rapolu, davem,
edumazet, haijun.liu, johannes, linuxwwan, m.chetan.kumar,
netdev, pabeni, ricardo.martinez, ryazanov.s.a
Le sam. 21 mai 2022 à 03:01, Jakub Kicinski <kuba@kernel.org> a écrit :
>
> On Fri, 20 May 2022 17:25:56 -0700 Jakub Kicinski wrote:
> > On Thu, 19 May 2022 09:29:12 +0200 Loic Poulain wrote:
> > > On Thu, 19 May 2022 at 09:26, Ziyang Xuan <william.xuanziyang@huawei.com> wrote:
> > > >
> > > > t7xx_cldma_clear_rxq() call t7xx_cldma_alloc_and_map_skb() in spin_lock
> > > > context, But __dev_alloc_skb() in t7xx_cldma_alloc_and_map_skb() uses
> > > > GFP_KERNEL, that will introduce scheduling factor in spin_lock context.
> > > >
> > > > Because t7xx_cldma_clear_rxq() is called after stopping CLDMA, so we can
> > > > remove the spin_lock from t7xx_cldma_clear_rxq().
> > > >
> > > > Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface")
> > > > Suggested-by: Ricardo Martinez <ricardo.martinez@linux.intel.com>
> > > > Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
> > >
> > > Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
> >
> > Wait, you reviewed two different fixes for the same issue?
> > Please say something when that happens I thought both are needed :/
Right, I've actually overlooked that the other patch has only one
atomic user, which becomes useless with this change.
>
>
> FWIW I pushed out the other one before I realized (they both apply
> without conflicts so I thought they fixed different issues)
> If this one is preferred please respin and squash a revert of
>
>
> 9ee152ee3ee3 into it.
Yes this one is preferred, I'll respin it. Sorry for this.
Thanks,
Loic
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2022-05-21 17:27 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-19 7:43 [PATCH net-next v3] net: wwan: t7xx: fix GFP_KERNEL usage in spin_lock context Ziyang Xuan
2022-05-19 7:29 ` Loic Poulain
2022-05-21 0:25 ` Jakub Kicinski
2022-05-21 1:01 ` Jakub Kicinski
2022-05-21 17:26 ` Loic Poulain
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.