From: Fabio Estevam
Date: Fri, 7 Nov 2014 10:26:15 -0200
> On Fri, Nov 7, 2014 at 7:02 AM, Lothar Waßmann
> wrote:
>> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
>> performance")
>> introduced a regression for i.MX28. The swap_buffer() function doing
>> the endian conv
On Fri, Nov 7, 2014 at 7:02 AM, Lothar Waßmann wrote:
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyond the
commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
performance")
introduced a regression for i.MX28. The swap_buffer() function doing
the endian conversion of the received data on i.MX28 may access memory
beyond the actual packet size in the DMA buffer. fec_enet_copybreak()
does
Changes wrt. v1:
- added some cleanup patches
- simplify handling of 'quirks' flags as suggested by Russell King.
- remove DIV_ROUND_UP() from byte swapping loop as suggested by
Eric Dumazet
Changes wrt. v2:
- rebased against next-20141028
- added some more cleanups in fec.h
- removed unused ret
From: Lothar Waßmann
Date: Wed, 5 Nov 2014 06:21:48 +0100
> David Miller wrote:
>> Because it goes "(n + (d - 1)) / d"
>>
>> Which for 'd' of 4 gives:
>>
>> 1 --> 4
>> 2 --> 4
>> 3 --> 4
>> 4 --> 4
>>
> '1', not '4'.
> The loop has to be done once for each (probably partial) WORD of input
> dat
Hi,
David Miller wrote:
> From: Lothar Waßmann
> Date: Tue, 4 Nov 2014 11:29:12 +0100
>
> > Hi David,
> >
> > Lothar Waßmann wrote:
> >> David Miller wrote:
> >> > From: Lothar Waßmann
> >> > Date: Thu, 30 Oct 2014 07:51:04 +0100
> >> >
> >> > >> Also, I don't thnk your DIV_ROUND_UP() elimina
From: Lothar Waßmann
Date: Tue, 4 Nov 2014 11:29:12 +0100
> Hi David,
>
> Lothar Waßmann wrote:
>> David Miller wrote:
>> > From: Lothar Waßmann
>> > Date: Thu, 30 Oct 2014 07:51:04 +0100
>> >
>> > >> Also, I don't thnk your DIV_ROUND_UP() eliminate for the loop
>> > >> in swap_buffer() is val
Hi David,
Lothar Waßmann wrote:
> David Miller wrote:
> > From: Lothar Waßmann
> > Date: Thu, 30 Oct 2014 07:51:04 +0100
> >
> > >> Also, I don't thnk your DIV_ROUND_UP() eliminate for the loop
> > >> in swap_buffer() is valid. The whole point is that the current
> > >> code handles buffers whi
Hi Lothar,
> Lothar Waßmann hat am 28. Oktober 2014 um 14:23
> geschrieben:
>
>
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may
Hi,
David Miller wrote:
> From: Lothar Waßmann
> Date: Thu, 30 Oct 2014 07:51:04 +0100
>
> >> Also, I don't thnk your DIV_ROUND_UP() eliminate for the loop
> >> in swap_buffer() is valid. The whole point is that the current
> >> code handles buffers which have a length which is not a multiple
>
From: Lothar Waßmann
Date: Thu, 30 Oct 2014 07:51:04 +0100
>> Also, I don't thnk your DIV_ROUND_UP() eliminate for the loop
>> in swap_buffer() is valid. The whole point is that the current
>> code handles buffers which have a length which is not a multiple
>> of 4 properly, after your change it
Hi,
David Miller wrote:
> From: Lothar Waßmann
> Date: Tue, 28 Oct 2014 14:22:55 +0100
>
> > Changes wrt. v1:
> > - added some cleanup patches
> > - simplify handling of 'quirks' flags as suggested by Russell King.
> > - remove DIV_ROUND_UP() from byte swapping loop as suggested by
> > Eric Du
From: Lothar Waßmann
Date: Tue, 28 Oct 2014 14:22:55 +0100
> Changes wrt. v1:
> - added some cleanup patches
> - simplify handling of 'quirks' flags as suggested by Russell King.
> - remove DIV_ROUND_UP() from byte swapping loop as suggested by
> Eric Dumazet
>
> Changes wrt. v2:
> - rebased a
Hi,
David Laight wrote:
> From: Lothar Waßmann
> > David Laight wrote:
> > > From: Lothar Waßmann
> > > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > > performance")
> > > > introduced a regression for i.MX28. The swap_buffer() function doing
> > > > the endian conv
commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
performance")
introduced a regression for i.MX28. The swap_buffer() function doing
the endian conversion of the received data on i.MX28 may access memory
beyond the actual packet size in the DMA buffer. fec_enet_copybreak()
does
Changes wrt. v1:
- added some cleanup patches
- simplify handling of 'quirks' flags as suggested by Russell King.
- remove DIV_ROUND_UP() from byte swapping loop as suggested by
Eric Dumazet
Changes wrt. v2:
- rebased against next-20141028
- added some more cleanups in fec.h
- removed unused ret
From: Lothar Waßmann
> David Laight wrote:
> > From: Lothar Waßmann
> > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > performance")
> > > introduced a regression for i.MX28. The swap_buffer() function doing
> > > the endian conversion of the received data on i.MX28 m
Hi,
David Laight wrote:
> From: Lothar Waßmann
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received data on i.MX28 may access memory
> > beyond
Hi,
David Laight wrote:
> From: Lothar Waßmann
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received data on i.MX28 may access memory
> > beyond
From: Lothar Waßmann
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyond the actual packet size in the DMA bu
From: Lothar Waßmann
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyond the actual packet size in the DMA buf
Changes wrt. v1:
- added some cleanup patches
- simplify handling of 'quirks' flags as suggested by Russell King.
- remove DIV_ROUND_UP() from byte swapping loop as suggested by
Eric Dumazet
Subject:
In-Reply-To:
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
th
commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
performance")
introduced a regression for i.MX28. The swap_buffer() function doing
the endian conversion of the received data on i.MX28 may access memory
beyond the actual packet size in the DMA buffer. fec_enet_copybreak()
does
On Wed, Oct 08, 2014 at 08:54:58AM +, David Laight wrote:
> Hmmm... in that case you may not want the compiler to convert the bit value
> to a 'bool' at all.
>
> Passing 'id_entry->driver_data' through (that doesn't look like a field name
> for
> 'quirk flags) would generate better code.
>
>
Hi,
David Laight wrote:
> From: Lothar Waßmann
> > David Laight wrote:
> > > From: Eric Dumazet
> > > > On Tue, 2014-10-07 at 15:19 +0200, Lothar Wamann wrote:
> > > > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > > > performance")
> > > > > introduced a regression
From: Sergei Shtylyov
> On 10/07/2014 05:19 PM, Lothar Wamann wrote:
>
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received data on i.MX28 may a
From: Lothar Waßmann
> David Laight wrote:
> > From: Eric Dumazet
> > > On Tue, 2014-10-07 at 15:19 +0200, Lothar Wamann wrote:
> > > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > > performance")
> > > > introduced a regression for i.MX28. The swap_buffer() function
Hi,
David Laight wrote:
> From: Eric Dumazet
> > On Tue, 2014-10-07 at 15:19 +0200, Lothar Wamann wrote:
> > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > performance")
> > > introduced a regression for i.MX28. The swap_buffer() function doing
> > > the endian conve
Hi,
David Laight wrote:
> From: Lothar
> > David Laight wrote:
> > > From: Lothar Waßmann
> > > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > > performance")
> > > > introduced a regression for i.MX28. The swap_buffer() function doing
> > > > the endian conversion o
Hi,
David Miller wrote:
> From: Lothar Waßmann
> Date: Tue, 7 Oct 2014 15:19:37 +0200
>
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received
From: Lothar Waßmann
Date: Tue, 7 Oct 2014 15:19:37 +0200
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyo
Hello.
On 10/07/2014 05:19 PM, Lothar Waßmann wrote:
commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
performance")
introduced a regression for i.MX28. The swap_buffer() function doing
the endian conversion of the received data on i.MX28 may access memory
beyond the actual
From: Eric Dumazet
> On Tue, 2014-10-07 at 15:19 +0200, Lothar Wamann wrote:
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received data on i.MX28
On Tue, 2014-10-07 at 14:23 +, David Laight wrote:
> The point I was making is that if you have to do a read-write of the received
> data (to byteswap it) then you might as well always copy it into a new skb
> that
> is just big enough for the actual receive frame.
+1
--
To unsubscribe fro
On Tue, 2014-10-07 at 15:19 +0200, Lothar Waßmann wrote:
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyond t
From: Lothar
> David Laight wrote:
> > From: Lothar Waßmann
> > > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > > performance")
> > > introduced a regression for i.MX28. The swap_buffer() function doing
> > > the endian conversion of the received data on i.MX28 may acces
Hi,
David Laight wrote:
> From: Lothar Waßmann
> > commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> > performance")
> > introduced a regression for i.MX28. The swap_buffer() function doing
> > the endian conversion of the received data on i.MX28 may access memory
> > beyond
Hi Lothar,
On Tue, Oct 7, 2014 at 10:19 AM, Lothar Waßmann
wrote:
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memor
From: Lothar Waßmann
> commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
> performance")
> introduced a regression for i.MX28. The swap_buffer() function doing
> the endian conversion of the received data on i.MX28 may access memory
> beyond the actual packet size in the DMA buf
commit 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx
performance")
introduced a regression for i.MX28. The swap_buffer() function doing
the endian conversion of the received data on i.MX28 may access memory
beyond the actual packet size in the DMA buffer. fec_enet_copybreak()
does
40 matches
Mail list logo
