generic-library/openssl
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

sparcv9_modes.pl: membars are reported as must-have.

Browse Source
This commit is contained in:
Andy Polyakov 2012-10-15 14:04:52 +00:00
parent d17b59e49f
commit fd3b0eb01d
1 changed files with 10 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
crypto/perlasm/sparcv9_modes.pl
View File

@ -12,14 +12,9 @@
# This is "cooperative" optimization, as it reduces overall pressure # This is "cooperative" optimization, as it reduces overall pressure
# on memory interface. Benefits can't be observed/quantified with # on memory interface. Benefits can't be observed/quantified with
# usual benchmarks, on the contrary you can notice that single-thread # usual benchmarks, on the contrary you can notice that single-thread
# performance for parallelizable modes is ~1.5% worse. Special note # performance for parallelizable modes is ~1.5% worse for largest
# about commented 'membar' instructions, otherwise recommended by # block sizes [though few percent better for not so long ones]. All
# manual. Rationale is following. Memory view is consistent from # this based on suggestions from David Miller.
# viewpoint of processor executing the code even when ASI in question
# is used. If thread on another processor has to access the result,
# its availability would have to be mediated and it can be done only
# through a syncronization operation which would requre ... 'membar'.
# All this based on suggestions from David Miller.
my ($inp,$out,$len,$key,$ivec,$enc)=map("%i$_",(0..5)); my ($inp,$out,$len,$key,$ivec,$enc)=map("%i$_",(0..5));
my ($ileft,$iright,$ooff,$omask,$ivoff,$blk_init)=map("%l$_",(0..7)); my ($ileft,$iright,$ooff,$omask,$ivoff,$blk_init)=map("%l$_",(0..7));
@ -213,7 +208,7 @@ $::code.=<<___;
brnz,pt $len, .L${bits}_cbc_enc_blk_loop brnz,pt $len, .L${bits}_cbc_enc_blk_loop
add $out, 8, $out add $out, 8, $out
!membar 0x0f membar #StoreLoad|#StoreStore
brnz,pt $blk_init, .L${bits}_cbc_enc_loop brnz,pt $blk_init, .L${bits}_cbc_enc_loop
mov $blk_init, $len mov $blk_init, $len
___ ___
@ -277,9 +272,9 @@ $::code.=<<___;
mov 0xff, $omask mov 0xff, $omask
sub $iright, $ileft, $iright sub $iright, $ileft, $iright
and $out, 7, $ooff and $out, 7, $ooff
cmp $len, 127 cmp $len, 255
movrnz $ooff, 0, $blk_init ! if ( $out&7 || movrnz $ooff, 0, $blk_init ! if ( $out&7 ||
movleu $::size_t_cc, 0, $blk_init ! $len<128 || movleu $::size_t_cc, 0, $blk_init ! $len<256 ||
brnz,pn $blk_init, .L${bits}cbc_dec_blk ! $inp==$out) brnz,pn $blk_init, .L${bits}cbc_dec_blk ! $inp==$out)
srl $omask, $ooff, $omask srl $omask, $ooff, $omask
@ -569,7 +564,7 @@ $::code.=<<___;
add $blk_init, $len, $len add $blk_init, $len, $len
andcc $len, 1, %g0 ! is number of blocks even? andcc $len, 1, %g0 ! is number of blocks even?
!membar 0x0f membar #StoreLoad|#StoreStore
bnz,pt %icc, .L${bits}_cbc_dec_loop bnz,pt %icc, .L${bits}_cbc_dec_loop
srl $len, 0, $len srl $len, 0, $len
brnz,pn $len, .L${bits}_cbc_dec_loop2x brnz,pn $len, .L${bits}_cbc_dec_loop2x
@ -630,9 +625,9 @@ ${alg}${bits}_t4_ctr32_encrypt:
mov 0xff, $omask mov 0xff, $omask
sub $iright, $ileft, $iright sub $iright, $ileft, $iright
and $out, 7, $ooff and $out, 7, $ooff
cmp $len, 127 cmp $len, 255
movrnz $ooff, 0, $blk_init ! if ( $out&7 || movrnz $ooff, 0, $blk_init ! if ( $out&7 ||
movleu $::size_t_cc, 0, $blk_init ! $len<128 || movleu $::size_t_cc, 0, $blk_init ! $len<256 ||
brnz,pn $blk_init, .L${bits}_ctr32_blk ! $inp==$out) brnz,pn $blk_init, .L${bits}_ctr32_blk ! $inp==$out)
srl $omask, $ooff, $omask srl $omask, $ooff, $omask
@ -884,7 +879,7 @@ $::code.=<<___;
add $blk_init, $len, $len add $blk_init, $len, $len
andcc $len, 1, %g0 ! is number of blocks even? andcc $len, 1, %g0 ! is number of blocks even?
!membar 0x0f membar #StoreLoad|#StoreStore
bnz,pt %icc, .L${bits}_ctr32_loop bnz,pt %icc, .L${bits}_ctr32_loop
srl $len, 0, $len srl $len, 0, $len
brnz,pn $len, .L${bits}_ctr32_loop2x brnz,pn $len, .L${bits}_ctr32_loop2x