diff mbox series

net: neterion: vxge: reduce stack usage in VXGE_COMPLETE_VPATH_TX

Message ID 20200716173247.78912-1-cuibixuan@huawei.com
State Changes Requested
Delegated to: David Miller
Headers show
Series net: neterion: vxge: reduce stack usage in VXGE_COMPLETE_VPATH_TX | expand

Commit Message

Bixuan Cui July 16, 2020, 5:32 p.m. UTC 
Fix the warning: [-Werror=-Wframe-larger-than=]

drivers/net/ethernet/neterion/vxge/vxge-main.c:
In function'VXGE_COMPLETE_VPATH_TX.isra.37':
drivers/net/ethernet/neterion/vxge/vxge-main.c:119:1:
warning: the frame size of 1056 bytes is larger than 1024 bytes

Signed-off-by: Bixuan Cui <cuibixuan@huawei.com>
---
 drivers/net/ethernet/neterion/vxge/vxge-main.c | 8 ++++++++
 1 file changed, 8 insertions(+)

--
2.17.1

Comments

Joe Perches July 16, 2020, 9:46 a.m. UTC | #1 
On Thu, 2020-07-16 at 17:32 +0000, Bixuan Cui wrote:
> Fix the warning: [-Werror=-Wframe-larger-than=]
[]
> diff --git a/drivers/net/ethernet/neterion/vxge/vxge-main.c b/drivers/net/ethernet/neterion/vxge/vxge-main.c
[]
> @@ -100,8 +100,14 @@ static inline void VXGE_COMPLETE_VPATH_TX(struct vxge_fifo *fifo)
>  	struct sk_buff **temp;
>  #define NR_SKB_COMPLETED 128
>  	struct sk_buff *completed[NR_SKB_COMPLETED];
> +	struct sk_buff **completed;
>  	int more;
> 
> +	completed = kcalloc(NR_SKB_COMPLETED, sizeof(*completed),
> +			    GFP_KERNEL);

I doubt this is a good idea.
Check the callers interrupt status.
Bixuan Cui July 16, 2020, 9:59 a.m. UTC | #2 
On 2020/7/16 17:46, Joe Perches wrote:
> I doubt this is a good idea.
> Check the callers interrupt status.
yes, it's not good idea to alloc memory in interrupt handler,
I will think more while fix warning. :)
Thanks.
kernel test robot July 16, 2020, 11:20 a.m. UTC | #3 
Hi Bixuan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on sparc-next/master]
[also build test WARNING on net-next/master net/master linus/master v5.8-rc5 next-20200716]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Bixuan-Cui/net-neterion-vxge-reduce-stack-usage-in-VXGE_COMPLETE_VPATH_TX/20200716-173219
base:   https://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-next.git master
config: s390-allyesconfig (attached as .config)
compiler: s390-linux-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=s390 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All warnings (new ones prefixed by >>):

      91 | #define this_cpu_add_8(pcp, val) arch_this_cpu_add(pcp, val, "laag", "agsi", long)
         |                                  ^~~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:380:11: note: in expansion of macro 'this_cpu_add_8'
     380 |   case 8: stem##8(variable, __VA_ARGS__);break;  \
         |           ^~~~
   include/linux/percpu-defs.h:509:33: note: in expansion of macro '__pcpu_size_call'
     509 | #define this_cpu_add(pcp, val)  __pcpu_size_call(this_cpu_add_, pcp, val)
         |                                 ^~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:519:33: note: in expansion of macro 'this_cpu_add'
     519 | #define this_cpu_sub(pcp, val)  this_cpu_add(pcp, -(typeof(pcp))(val))
         |                                 ^~~~~~~~~~~~
   include/linux/percpu-defs.h:521:28: note: in expansion of macro 'this_cpu_sub'
     521 | #define this_cpu_dec(pcp)  this_cpu_sub(pcp, 1)
         |                            ^~~~~~~~~~~~
   include/net/sch_generic.h:872:2: note: in expansion of macro 'this_cpu_dec'
     872 |  this_cpu_dec(sch->cpu_qstats->qlen);
         |  ^~~~~~~~~~~~
   include/net/sch_generic.h: In function 'qdisc_qstats_cpu_requeues_inc':
   arch/s390/include/asm/percpu.h:74:21: warning: comparison is always true due to limited range of data type [-Wtype-limits]
      74 |      ((szcast)val__ > -129) && ((szcast)val__ < 128)) {  \
         |                     ^
   arch/s390/include/asm/percpu.h:91:34: note: in expansion of macro 'arch_this_cpu_add'
      91 | #define this_cpu_add_8(pcp, val) arch_this_cpu_add(pcp, val, "laag", "agsi", long)
         |                                  ^~~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:380:11: note: in expansion of macro 'this_cpu_add_8'
     380 |   case 8: stem##8(variable, __VA_ARGS__);break;  \
         |           ^~~~
   include/linux/percpu-defs.h:509:33: note: in expansion of macro '__pcpu_size_call'
     509 | #define this_cpu_add(pcp, val)  __pcpu_size_call(this_cpu_add_, pcp, val)
         |                                 ^~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:520:28: note: in expansion of macro 'this_cpu_add'
     520 | #define this_cpu_inc(pcp)  this_cpu_add(pcp, 1)
         |                            ^~~~~~~~~~~~
   include/net/sch_generic.h:877:2: note: in expansion of macro 'this_cpu_inc'
     877 |  this_cpu_inc(sch->cpu_qstats->requeues);
         |  ^~~~~~~~~~~~
   include/net/sch_generic.h: In function 'qdisc_qstats_cpu_drop':
   arch/s390/include/asm/percpu.h:74:21: warning: comparison is always true due to limited range of data type [-Wtype-limits]
      74 |      ((szcast)val__ > -129) && ((szcast)val__ < 128)) {  \
         |                     ^
   arch/s390/include/asm/percpu.h:91:34: note: in expansion of macro 'arch_this_cpu_add'
      91 | #define this_cpu_add_8(pcp, val) arch_this_cpu_add(pcp, val, "laag", "agsi", long)
         |                                  ^~~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:380:11: note: in expansion of macro 'this_cpu_add_8'
     380 |   case 8: stem##8(variable, __VA_ARGS__);break;  \
         |           ^~~~
   include/linux/percpu-defs.h:509:33: note: in expansion of macro '__pcpu_size_call'
     509 | #define this_cpu_add(pcp, val)  __pcpu_size_call(this_cpu_add_, pcp, val)
         |                                 ^~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:520:28: note: in expansion of macro 'this_cpu_add'
     520 | #define this_cpu_inc(pcp)  this_cpu_add(pcp, 1)
         |                            ^~~~~~~~~~~~
   include/net/sch_generic.h:902:2: note: in expansion of macro 'this_cpu_inc'
     902 |  this_cpu_inc(sch->cpu_qstats->drops);
         |  ^~~~~~~~~~~~
   include/net/sch_generic.h: In function 'qdisc_update_stats_at_enqueue':
   arch/s390/include/asm/percpu.h:74:21: warning: comparison is always true due to limited range of data type [-Wtype-limits]
      74 |      ((szcast)val__ > -129) && ((szcast)val__ < 128)) {  \
         |                     ^
   arch/s390/include/asm/percpu.h:91:34: note: in expansion of macro 'arch_this_cpu_add'
      91 | #define this_cpu_add_8(pcp, val) arch_this_cpu_add(pcp, val, "laag", "agsi", long)
         |                                  ^~~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:380:11: note: in expansion of macro 'this_cpu_add_8'
     380 |   case 8: stem##8(variable, __VA_ARGS__);break;  \
         |           ^~~~
   include/linux/percpu-defs.h:509:33: note: in expansion of macro '__pcpu_size_call'
     509 | #define this_cpu_add(pcp, val)  __pcpu_size_call(this_cpu_add_, pcp, val)
         |                                 ^~~~~~~~~~~~~~~~
   include/net/sch_generic.h:1101:3: note: in expansion of macro 'this_cpu_add'
    1101 |   this_cpu_add(sch->cpu_qstats->backlog, pkt_len);
         |   ^~~~~~~~~~~~
   include/net/sch_generic.h: In function 'mini_qdisc_qstats_cpu_drop':
   arch/s390/include/asm/percpu.h:74:21: warning: comparison is always true due to limited range of data type [-Wtype-limits]
      74 |      ((szcast)val__ > -129) && ((szcast)val__ < 128)) {  \
         |                     ^
   arch/s390/include/asm/percpu.h:91:34: note: in expansion of macro 'arch_this_cpu_add'
      91 | #define this_cpu_add_8(pcp, val) arch_this_cpu_add(pcp, val, "laag", "agsi", long)
         |                                  ^~~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:380:11: note: in expansion of macro 'this_cpu_add_8'
     380 |   case 8: stem##8(variable, __VA_ARGS__);break;  \
         |           ^~~~
   include/linux/percpu-defs.h:509:33: note: in expansion of macro '__pcpu_size_call'
     509 | #define this_cpu_add(pcp, val)  __pcpu_size_call(this_cpu_add_, pcp, val)
         |                                 ^~~~~~~~~~~~~~~~
   include/linux/percpu-defs.h:520:28: note: in expansion of macro 'this_cpu_add'
     520 | #define this_cpu_inc(pcp)  this_cpu_add(pcp, 1)
         |                            ^~~~~~~~~~~~
   include/net/sch_generic.h:1270:2: note: in expansion of macro 'this_cpu_inc'
    1270 |  this_cpu_inc(miniq->cpu_qstats->drops);
         |  ^~~~~~~~~~~~
   drivers/net/ethernet/neterion/vxge/vxge-main.c: In function 'VXGE_COMPLETE_VPATH_TX':
   drivers/net/ethernet/neterion/vxge/vxge-main.c:103:19: error: conflicting types for 'completed'
     103 |  struct sk_buff **completed;
         |                   ^~~~~~~~~
   drivers/net/ethernet/neterion/vxge/vxge-main.c:102:18: note: previous declaration of 'completed' was here
     102 |  struct sk_buff *completed[NR_SKB_COMPLETED];
         |                  ^~~~~~~~~
   drivers/net/ethernet/neterion/vxge/vxge-main.c:126:2: error: implicit declaration of function 'free' [-Werror=implicit-function-declaration]
     126 |  free(completed);
         |  ^~~~
>> drivers/net/ethernet/neterion/vxge/vxge-main.c:126:2: warning: incompatible implicit declaration of built-in function 'free'
   drivers/net/ethernet/neterion/vxge/vxge-main.c:60:1: note: include '<stdlib.h>' or provide a declaration of 'free'
      59 | #include "vxge-main.h"
     +++ |+#include <stdlib.h>
      60 | #include "vxge-reg.h"
   cc1: some warnings being treated as errors

vim +/free +126 drivers/net/ethernet/neterion/vxge/vxge-main.c

    96	
    97	static inline void VXGE_COMPLETE_VPATH_TX(struct vxge_fifo *fifo)
    98	{
    99		struct sk_buff **skb_ptr = NULL;
   100		struct sk_buff **temp;
   101	#define NR_SKB_COMPLETED 128
   102		struct sk_buff *completed[NR_SKB_COMPLETED];
   103		struct sk_buff **completed;
   104		int more;
   105	
   106		completed = kcalloc(NR_SKB_COMPLETED, sizeof(*completed),
   107				    GFP_KERNEL);
   108		if (!completed)
   109			return;
   110	
   111		do {
   112			more = 0;
   113			skb_ptr = completed;
   114	
   115			if (__netif_tx_trylock(fifo->txq)) {
   116				vxge_hw_vpath_poll_tx(fifo->handle, &skb_ptr,
   117							NR_SKB_COMPLETED, &more);
   118				__netif_tx_unlock(fifo->txq);
   119			}
   120	
   121			/* free SKBs */
   122			for (temp = completed; temp != skb_ptr; temp++)
   123				dev_consume_skb_irq(*temp);
   124		} while (more);
   125	
 > 126		free(completed);
   127	}
   128	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
kernel test robot July 16, 2020, 2:11 p.m. UTC | #4 
Hi Bixuan,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on sparc-next/master]
[also build test ERROR on net-next/master net/master linus/master v5.8-rc5 next-20200716]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Bixuan-Cui/net-neterion-vxge-reduce-stack-usage-in-VXGE_COMPLETE_VPATH_TX/20200716-173219
base:   https://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-next.git master
config: arc-allyesconfig (attached as .config)
compiler: arc-elf-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   drivers/net/ethernet/neterion/vxge/vxge-main.c: In function 'VXGE_COMPLETE_VPATH_TX':
>> drivers/net/ethernet/neterion/vxge/vxge-main.c:103:19: error: conflicting types for 'completed'
     103 |  struct sk_buff **completed;
         |                   ^~~~~~~~~
   drivers/net/ethernet/neterion/vxge/vxge-main.c:102:18: note: previous declaration of 'completed' was here
     102 |  struct sk_buff *completed[NR_SKB_COMPLETED];
         |                  ^~~~~~~~~
>> drivers/net/ethernet/neterion/vxge/vxge-main.c:126:2: error: implicit declaration of function 'free' [-Werror=implicit-function-declaration]
     126 |  free(completed);
         |  ^~~~
   drivers/net/ethernet/neterion/vxge/vxge-main.c:60:1: note: 'free' is defined in header '<stdlib.h>'; did you forget to '#include <stdlib.h>'?
      59 | #include "vxge-main.h"
     +++ |+#include <stdlib.h>
      60 | #include "vxge-reg.h"
   cc1: some warnings being treated as errors

vim +/completed +103 drivers/net/ethernet/neterion/vxge/vxge-main.c

    96	
    97	static inline void VXGE_COMPLETE_VPATH_TX(struct vxge_fifo *fifo)
    98	{
    99		struct sk_buff **skb_ptr = NULL;
   100		struct sk_buff **temp;
   101	#define NR_SKB_COMPLETED 128
   102		struct sk_buff *completed[NR_SKB_COMPLETED];
 > 103		struct sk_buff **completed;
   104		int more;
   105	
   106		completed = kcalloc(NR_SKB_COMPLETED, sizeof(*completed),
   107				    GFP_KERNEL);
   108		if (!completed)
   109			return;
   110	
   111		do {
   112			more = 0;
   113			skb_ptr = completed;
   114	
   115			if (__netif_tx_trylock(fifo->txq)) {
   116				vxge_hw_vpath_poll_tx(fifo->handle, &skb_ptr,
   117							NR_SKB_COMPLETED, &more);
   118				__netif_tx_unlock(fifo->txq);
   119			}
   120	
   121			/* free SKBs */
   122			for (temp = completed; temp != skb_ptr; temp++)
   123				dev_consume_skb_irq(*temp);
   124		} while (more);
   125	
 > 126		free(completed);
   127	}
   128	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
Stephen Hemminger July 19, 2020, 5:05 p.m. UTC | #5 
On Thu, 16 Jul 2020 17:32:47 +0000
Bixuan Cui <cuibixuan@huawei.com> wrote:

> Fix the warning: [-Werror=-Wframe-larger-than=]
> 
> drivers/net/ethernet/neterion/vxge/vxge-main.c:
> In function'VXGE_COMPLETE_VPATH_TX.isra.37':
> drivers/net/ethernet/neterion/vxge/vxge-main.c:119:1:
> warning: the frame size of 1056 bytes is larger than 1024 bytes
> 
> Signed-off-by: Bixuan Cui <cuibixuan@huawei.com>

Dropping the NR_SKB_COMPLETED to 16 won't have much impact
on performance, and shrink the size.

Doing 16 skb's at a time instead of 128 probably costs
less than one allocation. Especially since it is unlikely
that the device completed that many transmits at once.
Bixuan Cui July 20, 2020, 1:57 a.m. UTC | #6 
On 2020/7/20 1:05, Stephen Hemminger wrote:
> On Thu, 16 Jul 2020 17:32:47 +0000
> Bixuan Cui <cuibixuan@huawei.com> wrote:
> 
>> Fix the warning: [-Werror=-Wframe-larger-than=]
>>
>> drivers/net/ethernet/neterion/vxge/vxge-main.c:
>> In function'VXGE_COMPLETE_VPATH_TX.isra.37':
>> drivers/net/ethernet/neterion/vxge/vxge-main.c:119:1:
>> warning: the frame size of 1056 bytes is larger than 1024 bytes
>>
>> Signed-off-by: Bixuan Cui <cuibixuan@huawei.com>
> Dropping the NR_SKB_COMPLETED to 16 won't have much impact
> on performance, and shrink the size.
> 
> Doing 16 skb's at a time instead of 128 probably costs
> less than one allocation. Especially since it is unlikely
> that the device completed that many transmits at once.
> 
> 
I will send the v2 patch based on your suggestions.
thanks
diff mbox series

Patch

diff --git a/drivers/net/ethernet/neterion/vxge/vxge-main.c b/drivers/net/ethernet/neterion/vxge/vxge-main.c
index b0faa737b817..97ddfc9debd4 100644
--- a/drivers/net/ethernet/neterion/vxge/vxge-main.c
+++ b/drivers/net/ethernet/neterion/vxge/vxge-main.c
@@ -100,8 +100,14 @@  static inline void VXGE_COMPLETE_VPATH_TX(struct vxge_fifo *fifo)
 	struct sk_buff **temp;
 #define NR_SKB_COMPLETED 128
 	struct sk_buff *completed[NR_SKB_COMPLETED];
+	struct sk_buff **completed;
 	int more;

+	completed = kcalloc(NR_SKB_COMPLETED, sizeof(*completed),
+			    GFP_KERNEL);
+	if (!completed)
+		return;
+
 	do {
 		more = 0;
 		skb_ptr = completed;
@@ -116,6 +122,8 @@  static inline void VXGE_COMPLETE_VPATH_TX(struct vxge_fifo *fifo)
 		for (temp = completed; temp != skb_ptr; temp++)
 			dev_consume_skb_irq(*temp);
 	} while (more);
+
+	free(completed);
 }

 static inline void VXGE_COMPLETE_ALL_TX(struct vxgedev *vdev)