* Re: [patch] alpha: fix alignment problem in csum_ipv6_magic()
[not found] <20070617012020.A7626@jurassic.park.msu.ru>
@ 2007-06-21 23:35 ` Andrew Morton
2007-06-22 11:26 ` Ivan Kokshaysky
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Morton @ 2007-06-21 23:35 UTC (permalink / raw)
To: Ivan Kokshaysky; +Cc: rth, linux-kernel, netdev, Dustin Marquess
> On Sun, 17 Jun 2007 01:20:20 +0400 Ivan Kokshaysky <ink@jurassic.park.msu.ru> wrote:
> Hopefully this fixes http://bugzilla.kernel.org/show_bug.cgi?id=8635
>
> The struct in6_addr passed to csum_ipv6_magic() is 4 byte aligned,
> so we can't use the regular 64-bit loads.
> Since the cost of handling of 4 byte and 1 byte aligned 64-bit data is
> roughly the same, this code can cope with any src/dst [mis]alignment.
>
> Signed-off-by: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
>
> Ivan.
>
> --- 2.6.22-rc4/arch/alpha/lib/ev6-csum_ipv6_magic.S Sun Feb 4 21:44:54 2007
> +++ linux/arch/alpha/lib/ev6-csum_ipv6_magic.S Sun Jun 17 00:41:53 2007
> @@ -46,6 +46,10 @@
> * add the 3 low ushorts together, generating a uint
> * a final add of the 2 lower ushorts
> * truncating the result.
> + *
> + * Misalignment handling added by Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> + * The cost is 16 instructions (~8 cycles), including two extra loads which
> + * may cause additional delay in rare cases (load-load replay traps).
> */
>
> .globl csum_ipv6_magic
> @@ -55,25 +59,45 @@
> csum_ipv6_magic:
> .prologue 0
>
> - ldq $0,0($16) # L : Latency: 3
> + ldq_u $0,0($16) # L : Latency: 3
> inslh $18,7,$4 # U : 0000000000AABBCC
> - ldq $1,8($16) # L : Latency: 3
> + ldq_u $1,8($16) # L : Latency: 3
> sll $19,8,$7 # U : U L U L : 0x00000000 00aabb00
>
> + and $16,7,$6 # E : src misalignment
> + ldq_u $5,15($16) # L : Latency: 3
> zapnot $20,15,$20 # U : zero extend incoming csum
> - ldq $2,0($17) # L : Latency: 3
> - sll $19,24,$19 # U : U L L U : 0x000000aa bb000000
> - inswl $18,3,$18 # U : 000000CCDD000000
> + ldq_u $2,0($17) # L : U L U L : Latency: 3
> +
> + extql $0,$6,$0 # U :
> + extqh $1,$6,$22 # U :
> + ldq_u $3,8($17) # L : Latency: 3
> + sll $19,24,$19 # U : U U L U : 0x000000aa bb000000
> +
> + cmoveq $6,$31,$22 # E : src aligned?
> + ldq_u $23,15($17) # L : Latency: 3
> + or $18,$4,$18 # E : 000000CCDDAABBCC
> + extql $1,$6,$1 # U : U L L U :
>
> - ldq $3,8($17) # L : Latency: 3
> - bis $18,$4,$18 # E : 000000CCDDAABBCC
> + or $0,$22,$0 # E : 1st src word complete
> + extqh $5,$6,$5 # U :
> addl $19,$7,$19 # E : <sign bits>bbaabb00
> - nop # E : U L U L
> + and $17,7,$6 # E : L U L U : dst misalignment
>
> + inswl $18,3,$18 # U : 000000CCDD000000
> + or $1,$5,$1 # E : 2nd src word complete
> + extql $2,$6,$2 # U :
> + extqh $3,$6,$22 # U : U L U U :
> +
> + cmoveq $6,$31,$22 # E : dst aligned?
> + extql $3,$6,$3 # U :
> addq $20,$0,$20 # E : begin summing the words
> + extqh $23,$6,$23 # U : L U L U :
> +
> srl $18,16,$4 # U : 0000000000CCDDAA
> + or $2,$22,$2 # E : 1st dst word complete
> zap $19,0x3,$19 # U : <sign bits>bbaa0000
> - nop # E : L U U L
> + or $3,$23,$3 # E : U L U L : 2nd dst word complete
>
> cmpult $20,$0,$0 # E :
> addq $20,$1,$20 # E :
> --- 2.6.22-rc4/arch/alpha/lib/csum_ipv6_magic.S Sun Feb 4 21:44:54 2007
> +++ linux/arch/alpha/lib/csum_ipv6_magic.S Sun Jun 17 00:29:28 2007
> @@ -7,6 +7,9 @@
> * __u32 len,
> * unsigned short proto,
> * unsigned int csum);
> + *
> + * Misalignment handling (which costs 16 instructions / 8 cycles)
> + * added by Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> */
>
> .globl csum_ipv6_magic
> @@ -16,37 +19,57 @@
> csum_ipv6_magic:
> .prologue 0
>
> - ldq $0,0($16) # e0 : load src & dst addr words
> + ldq_u $0,0($16) # e0 : load src & dst addr words
> zapnot $20,15,$20 # .. e1 : zero extend incoming csum
> extqh $18,1,$4 # e0 : byte swap len & proto while we wait
> - ldq $1,8($16) # .. e1 :
> + ldq_u $21,7($16) # .. e1 : handle misalignment
>
> extbl $18,1,$5 # e0 :
> - ldq $2,0($17) # .. e1 :
> + ldq_u $1,8($16) # .. e1 :
> extbl $18,2,$6 # e0 :
> - ldq $3,8($17) # .. e1 :
> + ldq_u $22,15($16) # .. e1 :
>
> extbl $18,3,$18 # e0 :
> + ldq_u $2,0($17) # .. e1 :
> sra $4,32,$4 # e0 :
> + ldq_u $23,7($17) # .. e1 :
> +
> + extql $0,$16,$0 # e0 :
> + ldq_u $3,8($17) # .. e1 :
> + extqh $21,$16,$21 # e0 :
> + ldq_u $24,15($17) # .. e1 :
> +
> sll $5,16,$5 # e0 :
> + or $0,$21,$0 # .. e1 : 1st src word complete
> + extql $1,$16,$1 # e0 :
> addq $20,$0,$20 # .. e1 : begin summing the words
>
> - sll $6,8,$6 # e0 :
> + extqh $22,$16,$22 # e0 :
> cmpult $20,$0,$0 # .. e1 :
> - extwh $19,7,$7 # e0 :
> - or $4,$18,$18 # .. e1 :
> + sll $6,8,$6 # e0 :
> + or $1,$22,$1 # .. e1 : 2nd src word complete
>
> - extbl $19,1,$19 # e0 :
> + extql $2,$17,$2 # e0 :
> + or $4,$18,$18 # .. e1 :
> + extqh $23,$17,$23 # e0 :
> or $5,$6,$5 # .. e1 :
> - or $18,$5,$18 # e0 : len complete
> - or $19,$7,$19 # .. e1 :
>
> - sll $19,48,$19 # e0 :
> + extql $3,$17,$3 # e0 :
> + or $2,$23,$2 # .. e1 : 1st dst word complete
> + extqh $24,$17,$24 # e0 :
> + or $18,$5,$18 # .. e1 : len complete
> +
> + extwh $19,7,$7 # e0 :
> + or $3,$24,$3 # .. e1 : 2nd dst word complete
> + extbl $19,1,$19 # e0 :
> addq $20,$1,$20 # .. e1 :
> - sra $19,32,$19 # e0 : proto complete
> +
> + or $19,$7,$19 # e0 :
> cmpult $20,$1,$1 # .. e1 :
> + sll $19,48,$19 # e0 :
> + nop # .. e0 :
>
> - nop # e0 :
> + sra $19,32,$19 # e0 : proto complete
> addq $20,$2,$20 # .. e1 :
> cmpult $20,$2,$2 # e0 :
> addq $20,$3,$20 # .. e1 :
> @@ -84,7 +107,7 @@ csum_ipv6_magic:
> extwl $0,2,$1 # e0 : fold 17-bit value
> zapnot $0,3,$0 # .. e1 :
> addq $0,$1,$0 # e0 :
> - not $0,$0 # e1 : and complement.
> + not $0,$0 # .. e1 : and complement.
>
> zapnot $0,3,$0 # e0 :
> ret # .. e1 :
In http://bugzilla.kernel.org/show_bug.cgi?id=8659, Dustin is reporting
that this patch broke tcp-on-ipv6.
^ permalink raw reply [flat|nested] 3+ messages in thread