diff mbox

[dpdk-dev,1/2] ixgbe: fix setup of mbuf initializer template

Message ID 1417703181-23093-2-git-send-email-jean-mickael.guerin@6wind.com (mailing list archive)
State Accepted, archived
Headers show

Commit Message

Jean-Mickael Guerin Dec. 4, 2014, 2:26 p.m. UTC
Add a compiler barrier to make sure all fields covered by
the marker rearm_data are assigned before the read.

Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
Acked-by: David Marchand <david.marchand@6wind.com>
Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
---
 lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 3 +++
 1 file changed, 3 insertions(+)

Comments

Bruce Richardson Dec. 4, 2014, 2:39 p.m. UTC | #1
On Thu, Dec 04, 2014 at 03:26:20PM +0100, Jean-Mickael Guerin wrote:
> Add a compiler barrier to make sure all fields covered by
> the marker rearm_data are assigned before the read.
> 
> Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
> Acked-by: David Marchand <david.marchand@6wind.com>
> Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")

Acked-by: Bruce Richardson <bruce.richardson@intel.com>

> ---
>  lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> index 579bc46..c1b5a78 100644
> --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> @@ -739,6 +739,9 @@ ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq)
>  	mb_def.buf_len = rxq->mb_pool->elt_size - sizeof(struct rte_mbuf);
>  	mb_def.port = rxq->port_id;
>  	rte_mbuf_refcnt_set(&mb_def, 1);
> +
> +	/* prevent compiler reordering: rearm_data covers previous fields */
> +	rte_compiler_barrier();
>  	rxq->mbuf_initializer = *((uint64_t *)&mb_def.rearm_data);
>  	return 0;
>  }
> -- 
> 2.1.3
>
Ananyev, Konstantin Dec. 4, 2014, 2:42 p.m. UTC | #2
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Bruce Richardson
> Sent: Thursday, December 04, 2014 2:40 PM
> To: Jean-Mickael Guerin
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/2] ixgbe: fix setup of mbuf initializer template
> 
> On Thu, Dec 04, 2014 at 03:26:20PM +0100, Jean-Mickael Guerin wrote:
> > Add a compiler barrier to make sure all fields covered by
> > the marker rearm_data are assigned before the read.
> >
> > Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
> > Acked-by: David Marchand <david.marchand@6wind.com>
> > Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
> 
> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
> 
> > ---
> >  lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 3 +++
> >  1 file changed, 3 insertions(+)
> >
> > diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> > index 579bc46..c1b5a78 100644
> > --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> > +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> > @@ -739,6 +739,9 @@ ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq)
> >  	mb_def.buf_len = rxq->mb_pool->elt_size - sizeof(struct rte_mbuf);
> >  	mb_def.port = rxq->port_id;
> >  	rte_mbuf_refcnt_set(&mb_def, 1);
> > +
> > +	/* prevent compiler reordering: rearm_data covers previous fields */
> > +	rte_compiler_barrier();
> >  	rxq->mbuf_initializer = *((uint64_t *)&mb_def.rearm_data);
> >  	return 0;
> >  }
> > --

Hmm, can someone explain to me why do we need a compiler barrier here?
Konstantin

> > 2.1.3
> >
Jean-Mickael Guerin Dec. 4, 2014, 3:15 p.m. UTC | #3
On 04/12/2014 15:42, Ananyev, Konstantin wrote:
>
>
>> -----Original Message-----
>> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Bruce Richardson
>> Sent: Thursday, December 04, 2014 2:40 PM
>> To: Jean-Mickael Guerin
>> Cc: dev@dpdk.org
>> Subject: Re: [dpdk-dev] [PATCH 1/2] ixgbe: fix setup of mbuf initializer template
>>
>> On Thu, Dec 04, 2014 at 03:26:20PM +0100, Jean-Mickael Guerin wrote:
>>> Add a compiler barrier to make sure all fields covered by
>>> the marker rearm_data are assigned before the read.
>>>
>>> Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
>>> Acked-by: David Marchand <david.marchand@6wind.com>
>>> Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
>>
>> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
>>
>>> ---
>>>   lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 3 +++
>>>   1 file changed, 3 insertions(+)
>>>
>>> diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
>>> index 579bc46..c1b5a78 100644
>>> --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
>>> +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
>>> @@ -739,6 +739,9 @@ ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq)
>>>   	mb_def.buf_len = rxq->mb_pool->elt_size - sizeof(struct rte_mbuf);
>>>   	mb_def.port = rxq->port_id;
>>>   	rte_mbuf_refcnt_set(&mb_def, 1);
>>> +
>>> +	/* prevent compiler reordering: rearm_data covers previous fields */
>>> +	rte_compiler_barrier();
>>>   	rxq->mbuf_initializer = *((uint64_t *)&mb_def.rearm_data);
>>>   	return 0;
>>>   }
>>> --
>
> Hmm, can someone explain to me why do we need a compiler barrier here?
> Konstantin

rearm_data is a separate field and as well an array of length zero, 
overlapping on purpose the fields data_off buf_len, port, refcnt.
It might depend on compiler, but I could see a wrong value of 0UL for 
mbuf_initializer without the barrier (gcc 4.4.6).


>
>>> 2.1.3
>>>
Ananyev, Konstantin Dec. 4, 2014, 4:22 p.m. UTC | #4
> -----Original Message-----
> From: Jean-Mickael Guerin [mailto:jean-mickael.guerin@6wind.com]
> Sent: Thursday, December 04, 2014 3:15 PM
> To: Ananyev, Konstantin; Richardson, Bruce
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/2] ixgbe: fix setup of mbuf initializer template
> 
> On 04/12/2014 15:42, Ananyev, Konstantin wrote:
> >
> >
> >> -----Original Message-----
> >> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Bruce Richardson
> >> Sent: Thursday, December 04, 2014 2:40 PM
> >> To: Jean-Mickael Guerin
> >> Cc: dev@dpdk.org
> >> Subject: Re: [dpdk-dev] [PATCH 1/2] ixgbe: fix setup of mbuf initializer template
> >>
> >> On Thu, Dec 04, 2014 at 03:26:20PM +0100, Jean-Mickael Guerin wrote:
> >>> Add a compiler barrier to make sure all fields covered by
> >>> the marker rearm_data are assigned before the read.
> >>>
> >>> Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
> >>> Acked-by: David Marchand <david.marchand@6wind.com>
> >>> Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
> >>
> >> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
> >>
> >>> ---
> >>>   lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c | 3 +++
> >>>   1 file changed, 3 insertions(+)
> >>>
> >>> diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> >>> index 579bc46..c1b5a78 100644
> >>> --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> >>> +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
> >>> @@ -739,6 +739,9 @@ ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq)
> >>>   	mb_def.buf_len = rxq->mb_pool->elt_size - sizeof(struct rte_mbuf);
> >>>   	mb_def.port = rxq->port_id;
> >>>   	rte_mbuf_refcnt_set(&mb_def, 1);
> >>> +
> >>> +	/* prevent compiler reordering: rearm_data covers previous fields */
> >>> +	rte_compiler_barrier();
> >>>   	rxq->mbuf_initializer = *((uint64_t *)&mb_def.rearm_data);
> >>>   	return 0;
> >>>   }
> >>> --
> >
> > Hmm, can someone explain to me why do we need a compiler barrier here?
> > Konstantin
> 
> rearm_data is a separate field and as well an array of length zero,
> overlapping on purpose the fields data_off buf_len, port, refcnt.
> It might depend on compiler, but I could see a wrong value of 0UL for
> mbuf_initializer without the barrier (gcc 4.4.6).

Ah ok then.
Probably it is some sort of bug in old version of the gcc.

> 
> 
> >
> >>> 2.1.3
> >>>
Thomas Monjalon Dec. 5, 2014, 10:07 p.m. UTC | #5
> > Add a compiler barrier to make sure all fields covered by
> > the marker rearm_data are assigned before the read.
> > 
> > Signed-off-by: Jean-Mickael Guerin <jean-mickael.guerin@6wind.com>
> > Acked-by: David Marchand <david.marchand@6wind.com>
> > Fixes: 0ff3324da2 ("ixgbe: rework vector pmd following mbuf changes")
> 
> Acked-by: Bruce Richardson <bruce.richardson@intel.com>

Applied

The patch 2/2 won't be applied because another alternative has been approved.

Thanks
diff mbox

Patch

diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
index 579bc46..c1b5a78 100644
--- a/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
+++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c
@@ -739,6 +739,9 @@  ixgbe_rxq_vec_setup(struct igb_rx_queue *rxq)
 	mb_def.buf_len = rxq->mb_pool->elt_size - sizeof(struct rte_mbuf);
 	mb_def.port = rxq->port_id;
 	rte_mbuf_refcnt_set(&mb_def, 1);
+
+	/* prevent compiler reordering: rearm_data covers previous fields */
+	rte_compiler_barrier();
 	rxq->mbuf_initializer = *((uint64_t *)&mb_def.rearm_data);
 	return 0;
 }