[dpdk-dev,v2] examples/l3fwd: fix aliasing in port grouping

Message ID 20171103104351.30181-1-gprathyusha@caviumnetworks.com (mailing list archive)
State Accepted, archived
Delegated to: Ferruh Yigit
Headers

Checks

Context Check Description
ci/checkpatch success coding style OK
ci/Intel-compilation success Compilation OK

Commit Message

Guduri Prathyusha Nov. 3, 2017, 10:43 a.m. UTC
  With -f-strict-aliasing enabled by default from -O2, gcc > 5.x gives
undefined behavior in port_groupx4 in ARM. 'pn' and 'pnum' are
two different pointers pointing to same chunk of memory and
with -f-strict-aliasing the pointers are assumed to be pointing to
different memory and compiler reorders instructions that depend on
pnum and pn. This breaks port grouping algorithm.

This patch eliminates the above problem by introducing a compiler
barrier between the instructions that depend on pnum, pn and lp.

Fixes: 569b290cdb36 ("examples/l3fwd: add NEON implementation")

Signed-off-by: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
---
 examples/l3fwd/l3fwd_neon.h | 1 +
 1 file changed, 1 insertion(+)
  

Comments

Jerin Jacob Nov. 6, 2017, 4:29 a.m. UTC | #1
-----Original Message-----
> Date: Fri,  3 Nov 2017 16:13:51 +0530
> From: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
> To: tomasz.kantecki@intel.com
> CC: Jianbo.Liu@arm.com, guduriprathyusha@gmail.com,
>  konstantin.ananyev@intel.com, dev@dpdk.org, Guduri Prathyusha
>  <gprathyusha@caviumnetworks.com>
> Subject: [dpdk-dev] [PATCH v2] examples/l3fwd: fix aliasing in port grouping
> X-Mailer: git-send-email 2.14.1
> 
> With -f-strict-aliasing enabled by default from -O2, gcc > 5.x gives
> undefined behavior in port_groupx4 in ARM. 'pn' and 'pnum' are
> two different pointers pointing to same chunk of memory and
> with -f-strict-aliasing the pointers are assumed to be pointing to
> different memory and compiler reorders instructions that depend on
> pnum and pn. This breaks port grouping algorithm.
> 
> This patch eliminates the above problem by introducing a compiler
> barrier between the instructions that depend on pnum, pn and lp.
> 
> Fixes: 569b290cdb36 ("examples/l3fwd: add NEON implementation")
> 
> Signed-off-by: Guduri Prathyusha <gprathyusha@caviumnetworks.com>

Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

> ---
>  examples/l3fwd/l3fwd_neon.h | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/examples/l3fwd/l3fwd_neon.h b/examples/l3fwd/l3fwd_neon.h
> index 4bc161394..b319b5a92 100644
> --- a/examples/l3fwd/l3fwd_neon.h
> +++ b/examples/l3fwd/l3fwd_neon.h
> @@ -114,6 +114,7 @@ port_groupx4(uint16_t pn[FWDSTEP + 1], uint16_t *lp, uint16x8_t dp1,
>  
>  	/* update last port counter. */
>  	lp[0] += gptbl[v].lpv;
> +	rte_compiler_barrier();
>  
>  	/* if dest port value has changed. */
>  	if (v != GRPMSK) {
> -- 
> 2.14.1
>
  
Jianbo Liu Nov. 6, 2017, 8:18 a.m. UTC | #2
The 11/06/2017 09:59, Jerin Jacob wrote:
> -----Original Message-----
> > Date: Fri,  3 Nov 2017 16:13:51 +0530
> > From: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
> > To: tomasz.kantecki@intel.com
> > CC: Jianbo.Liu@arm.com, guduriprathyusha@gmail.com,
> >  konstantin.ananyev@intel.com, dev@dpdk.org, Guduri Prathyusha
> >  <gprathyusha@caviumnetworks.com>
> > Subject: [dpdk-dev] [PATCH v2] examples/l3fwd: fix aliasing in port grouping
> > X-Mailer: git-send-email 2.14.1
> >
> > With -f-strict-aliasing enabled by default from -O2, gcc > 5.x gives
> > undefined behavior in port_groupx4 in ARM. 'pn' and 'pnum' are
> > two different pointers pointing to same chunk of memory and
> > with -f-strict-aliasing the pointers are assumed to be pointing to
> > different memory and compiler reorders instructions that depend on
> > pnum and pn. This breaks port grouping algorithm.
> >
> > This patch eliminates the above problem by introducing a compiler
> > barrier between the instructions that depend on pnum, pn and lp.
> >
> > Fixes: 569b290cdb36 ("examples/l3fwd: add NEON implementation")
> >
> > Signed-off-by: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
>
> Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
>

Acked-by: Jianbo Liu <jianbo.liu@arm.com>
IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
  
Ferruh Yigit Nov. 7, 2017, 8:12 a.m. UTC | #3
On 11/6/2017 12:18 AM, Jianbo Liu wrote:
> The 11/06/2017 09:59, Jerin Jacob wrote:
>> -----Original Message-----
>>> Date: Fri,  3 Nov 2017 16:13:51 +0530
>>> From: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
>>> To: tomasz.kantecki@intel.com
>>> CC: Jianbo.Liu@arm.com, guduriprathyusha@gmail.com,
>>>  konstantin.ananyev@intel.com, dev@dpdk.org, Guduri Prathyusha
>>>  <gprathyusha@caviumnetworks.com>
>>> Subject: [dpdk-dev] [PATCH v2] examples/l3fwd: fix aliasing in port grouping
>>> X-Mailer: git-send-email 2.14.1
>>>
>>> With -f-strict-aliasing enabled by default from -O2, gcc > 5.x gives
>>> undefined behavior in port_groupx4 in ARM. 'pn' and 'pnum' are
>>> two different pointers pointing to same chunk of memory and
>>> with -f-strict-aliasing the pointers are assumed to be pointing to
>>> different memory and compiler reorders instructions that depend on
>>> pnum and pn. This breaks port grouping algorithm.
>>>
>>> This patch eliminates the above problem by introducing a compiler
>>> barrier between the instructions that depend on pnum, pn and lp.
>>>
>>> Fixes: 569b290cdb36 ("examples/l3fwd: add NEON implementation")
>>>
>>> Signed-off-by: Guduri Prathyusha <gprathyusha@caviumnetworks.com>
>>
>> Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
> 
> Acked-by: Jianbo Liu <jianbo.liu@arm.com>

Applied to dpdk/master, thanks.
  

Patch

diff --git a/examples/l3fwd/l3fwd_neon.h b/examples/l3fwd/l3fwd_neon.h
index 4bc161394..b319b5a92 100644
--- a/examples/l3fwd/l3fwd_neon.h
+++ b/examples/l3fwd/l3fwd_neon.h
@@ -114,6 +114,7 @@  port_groupx4(uint16_t pn[FWDSTEP + 1], uint16_t *lp, uint16x8_t dp1,
 
 	/* update last port counter. */
 	lp[0] += gptbl[v].lpv;
+	rte_compiler_barrier();
 
 	/* if dest port value has changed. */
 	if (v != GRPMSK) {