From 01df6ed4a99f8164855c1ce049986e8b6137d7ff Mon Sep 17 00:00:00 2001 From: Evghenii Date: Wed, 13 Nov 2013 19:13:04 +0100 Subject: [PATCH] added ispc timers w/o task --- examples_cuda/sort/sort | Bin 58497 -> 58977 bytes examples_cuda/sort/sort.cpp | 24 ++++++++++++------------ examples_cuda/sort/sort_cu.cpp | 2 +- 3 files changed, 13 insertions(+), 13 deletions(-) diff --git a/examples_cuda/sort/sort b/examples_cuda/sort/sort index 64c32dd02ce9ef2b3774fb8a3734c07acd05e12d..8622b05dc978593bde5d9b014cd4680f9a0699af 100755 GIT binary patch delta 11189 zcmb7K3tUvy*5Bs01kaUhY6 zTDZ^opP|ih!0WVqcPW3;n!#M!h;du8%C|uC>oTc;Hsw@7urMoVEY|wf5R; zKh8RP&diDG(1o-(AO-Kj1ASSf{9XpddYUyi>>b((#9N ze29*3(D6nc-=yRFFrKe}vrZAJYoO@(M|Heg$D4Hg9UUL0X9UsMc3q};7Q^e~U4ASwEj*rywvW_3Dqu zcCfZ82KNEX*-3!iY6!fT_jhz7Is$O2n8_T+IwGL5!)y)xhw4#F#={PKIAajH$y_ z&hRwEm@!IpEm=> zl;CP$_&Q?D1TH7T7ZGCyaFsKB4zUSwDZ{4`haoOx*wTmu-s&zZBfdn8cepEq;e&|r z&UTp@u0)J?v@4C_9fiWu);R}{mW5aW&OGBW%&V!UBp{)hv^ijcrN)kT=W zn$7fbm%f(r^%9Zm4>*~qH+PeW@&)K`78c6cmwjcs&&CgpM9y_)ik=2`{}k#=#aejG zKQ9l;10iSMc|^ATbD=o3rUJTQjc!h@MwO@85Mo8H0+D8a;EtC}*L4lGFq3oK#F}^w z?nmt*3J> zv$y+vfwse&=ATnyVG`uIVkgv=9jYeff>WF;iVY1r1;q%{5F@nl0I;mvkc z(&GklDL}R_^;Nb5El+Qfv+vEw>Lk}En=k=blByx|kxT7m+Z0nqQZ=TYB-=l#x0Fa0 zl5D#Yb!2~lw)#&id8mfl~SlE_kQ+xQVWK@2X&HayRAHR&*K?mJA_q^{81(0D;u3^L*$K3w) z30vpAV?6b#3HP!tGDGDGq%@fE!Eo!y1z9FpF4$=*2RBMb=o3SR&ynl0VL-Ae4MnO+ zzHXT4(``8h`O}`+U~1pXV!3XZa9Fk%vv^>5sU~0LINIE!hln1TK0vw`2fmSp&%ju!Ul zZJ7;js`2o8!!Ys0eq87qhE5$!HXH+7vgi%@(U8+%?CSogHvd&36}X8~?dlRI+tnpc zhUvz2R(9Qi2peG+3d?)Q1y6XA?5V}xNeLZ_4VIPLyR30JzmiaR9B6oebZ z3Z}|j%}dL^GpZXT2sg-L@X9vG2C!|lLBd-$2&xR-|9FS=0YAUHL%d<+`Wedh=2>WAaf<$*)xedw?Cy&a`49jwhnpR&bl zL*N~4k5E&FE?wlC%FsnJ`iA`wk4_7b8GDP~4eI&icxcTQz7$B^!mI6C)*iKnoHKCc z$syUE!)}+P+wPY1aC^}oEdFU&0{!A217nsRgn#Q6-y5g)w-)~rSpD(hAF(i#|6cs< zyaTrQhoFO2>0z+8_&CJs^0tewJa?0&k9-r;8$oXfKTrYF3DcIIW5O*1m3+vou=q+} zu>Mo-`Apwb+XcnOz%2x~l`H_&PJ9|Vl{BcwIOP?G@CiCO>7}Ifq^U_$WZUJbHPwe8 zqwTMNOXUJ^y~m9S&J6a3kQ~5-yinb+*G!bx(Or0Futivz2aem}v)h$>(=nvia6vqM zozCf9Alh#_?)Ba;h;}!vGRBJ^UZ?*yo)+J{N>B6|*g5YSJ_--II>Ljtx(SA+{QU~` z5A81YxJn0xMh0BE;_=kXER^k|GXSrDr{!ONrdho(Ye$`@rmP-cP@d z>S0iuGkPA-v`_S-G7A79rZC!3Gt!d!?wjZWm~Om?*LD| zImdimR;K@NH5>=w2g@{lIXA99oCGhy7Ht|ylJlCkiT^4#(6Kq({C@Rz3LYz_wrG=J_-z2Wg) zf6LF(Q}ojCT@Qahg^=32bHirPPev3+{5ToVclVDRI=5Kra!v(}wXT^iBzm1yi8Ba% z&ePcm&kNgVbwW4cCweL&OfWlsO&H)MWYX|ayM+6WA4dHk3KHG-_*vl%`oR;Q8=T_^ zIrP8Ff|`Ze#dE_#>Ez@OB0_fl?L^reN$c( zuF~Zx-GqL$B&DxV>e!b89qy!xyi0h@@&1^@VqnQr&}_^2!?nMp9msV;M?S@l#i>h0 zzdRWfFi?M)-hX;)#KI>L;z%b-U6!bTk94BcYq3f|#}lQnEPCRZX%7!aDl2@@SvoRp znBa0`rajtD*g`kFI7-NIG`)C65RN!@PE7L>7CC&T+66;aJdXCwptCsIUO&(e(sx9B zH4J=Q!_h*g$5haWpw|fmu9*sG)lM7evX^EEc6#omL?O{(G#3cMT*tPTzwj2SY5L3y z!Yw-bl~KaCj#aPh5`-kj$k}7Ogxii)b54pv4V^SMT&SekbITuoxF7gkJhvl6o$;k# z9DJ)p^hjUoKd-Ul2>8|sogSz?d%xGyj;r$u#ejXGjB?iA%{ZKfe7<8{W|bJ=4r#+z zJJJ_!6vd8zrFR!UC%Tu=re6B2^b)#x=_>JW>*>I(-eOfTeJX2! zxVD%s%rc6v7t@VdQ$;eL-pM*BMvbHm)^7CVvbRKQ8mlf@K1D3dVAZS3SBNu;n2*eL zw8#=^N?r$dPQhMy$fyTbCT(yy+;0R%wp~<0*U|HqDd_%jYrg9KUCnjwigeBOFDsUb z?o8H8(aP|yW7lE=%7R0t#A~yZxV5xl<(uN#w;+IEFJ)5!osk_W&M%ezWQsVIDQD-Xn~n zKdg@G_s|;56FDYdJPyMdHyosL&8CtBN!yRgSsRVbONjOEgr}eD%DZ`>E*NQ2DB4YL zz`A_Um6kC#b zY&FfZz1!Zsifb>dMs53J8m?VUpIy^e=tUQ;**&d8B&kAA4dYVAW`C37QP9?QOfD!kMM2_Jf>?MypG46v`KESH!TT;{<|-=X zceXTxW`NTI^s{ zfkTbVJW@H43!&K)MIhvCIJM-y59ce_Znyw@ImSp&=C8F3KxaV8%*@7)hO1r1Ao)?TM=as*5xKNn9N} zTOFTg`vMJnh{fF#pSgk)TVz(5v)%=0Cr9Apr@u`d0W$Fm4z8l0+| zL7M5;A0>$HcGU0V1>z^|sO95zqBVpDY>gDtLg?dLM+cUMNYEd6hA^=B=4_|$Y<*oc zo2Xw|Z!yC}2Y|4e=-9H}V+&0ZGxBFwFfQA!5**_j;GHlprEToEn!vG$d;|J&6Wvjk z9%>yZF%6#h^kqDN@i1K(5j6Uf7XsZwB&IFJnStV40U>^p8(26@V%UtC1d8v`?KI}o zL4odgiQx=fAt=7zZ>I}Bg&`%-bDypf^62cptra%VyMKF8$e_=D_61yp&({isu7hhK z?X&GA!AMtZd!_rV^BxcP!5o?3x)b6uQ{juXe(;5c6_`lB@&PoX&fC)Pj>HTxAe_%x zn^lZXko$(~!3V?NbBwmAKhH4^6!B$_VdWYAg=1KGhW+1_pavs^57=9XkrS}=%pjg) zxVIP{&oSf|7T^qyAz3oKVY37^Lh#2649dwd{&s=*JjVu({kKR^gAENIZ~S(;h&zNw zCToz)4FWhW=K4U6%Q^1M@ga`8aNNW({&s-@WO0ApIJUkgL5+|g_;^GAG51U`soalH}h79`%~L?5Ul z_T~YEava3*qkskY4CUCw@koxtI8NrcFUP4I_v3gx$KkmC=x_=r`g4QBJfH!9F`!aD zf`J@MA4pJBe|2O5f2?Bs@d9x(*W&{n@kp*8%(1mpf*St&z@i#ta)Tj2VuLb{qd9i+ z1~D8PKa`*bGXWp8FXcFvV}-Ym0S+p7DI4m)se zlHf{60(=k~H`AXTZ;5#&bj{8;L}fSqdFTC3#u_uI9+YM()CJOJ!P{Z{d#NCn?V$&E zM~Y2*=yedjm2^@?KRTwuAWB&6Ycca%wfSv{`EYZ2lgnSYB;3xI%MM(2=CUW3CN3km z9LnWLE|a-TQ>BHB=fo5)XK*=(%M30vxy<6y%4H6h+CVg!--a*dcta~5Zl}^#1Z2$? z?09?6avw3zN`tE-{qjoG34+-lR~^@>l*7^zb?TjTWA$J^-QQ^%tK$3!)~>0 z!*1FLP~i}^NWj53KkFVBuxJyYTB}ztVj$W~wC4SZHW#h=PDGoO)_efbW~VhD2>)|3BQ;6S z6EFDVi8f(+g%{~Uw3%zocO}{sw&uGLykoj}1(60i1Y~!jpzIk;v}tQ?--DD^(Xn;u z(c1E9YRp^0Z-f=kcnpi4u1gQD*lT7jFa=^Mf{#FY5a(Vxu731b!q=JU;IfCQI}t12 zI~rfl`2Qa4N#I3zE~b^xbM=vy!@M_q@?yCI9|0DvnQvc>e+0ro9^bKI!RHp1&WsNx z_}e9xAa;cu$oVqP<8u(p)11e*yHK8JY48SDH1P&n2YJA=c=`D*3Y$SGV)i{q!(Q5W zI5Jq<9QZWC@)fFu08+M{1|NyEH19Vv8nYyp-f)t^NyXQetN_L`2zh8$!Dlm2rO9K= zKA7O!TrAo!(>d>3sq!IA3&#h}<2zU=lUo{W;0mctgB{32$47a`cyPgTfY}=ezTLwT z!LEeVb9*=vhJG0-k}Nt7gt?Ng{4&yFtuiw+ZL}#~>S#%g%4>`IJmVj1d?#QmX2u7T zoHl;uX@08B%nlZYm1P=#fbn$jSN%o*B0A}-NQ?IL(Yn0BOyM5l8Ax>ie-UJ{w%p7b zO@e=i9|XJryHLw>T7x*?jqu+ues9utMl(a2!@qXHHlI~nUkNPyUp zG_0eikEV;^HFU(+>G+oF?XUkL!T~?vSY&Y9)TTT{NGRMyg(f=b*ju7|2)%f0w3wDl zqmPdkOH1gITzquu#xcTv2*<1iE0wC#6>VQ|z90N#H&9nlbJX?ym!g1VC! zOzNhe0*JF86~;C$2=X^o?U zttM5iegAp}&g~dg;V^oz5q9%@dcAS97zImlGEy|IrLUfZ*|pIfCnLw2+cd3%I|T=- z{eM*3Xs3b4i7_XrT{ZAXegz(v+r5qs_$D&C?KiwX!#4y}6TVBzd9xM3^8u{+rUbr; rIzs=azaYjArGNU?B*qM-pMTq3h@l1F_7;c60tga^4t3OjTiNkH{;F6k delta 10640 zcmaJ{4_s8m_Mdxkbwz3Ury`23xG4Txi9ZS|cJYB;jf_l7(`-|pKc4xLka}9^t^(?^ zpxBUES&>V!gTSLW$i2+IPW$wE5{M9_;oDEl|-k#m+vWOX-ME610 z-p;!Ov9Z!MPrqJgZQZB#ha=$*)sFVx;bT;WT^lP^UI$3ayAR&s`!P%PV&wgF0f;m9 zm1vc>5>$vj7G#X3fuDv?)bO1&yrkiAP`)B-cot@^$k6azh}CCK$V`nwe}^U(4c|?} zXKMKF8a_+I2Wa?Z8vY(1kN#~Mg~6u)xlY3eYWNZjAEeNw`MQd`Ch0YWNlnAEDtD4c}kGw`zE! zhQFcV2Qc2Z{s$}!Vk9D2O}*$f{6GyKsNqc-K1{<$X?UZCAEe=* z4L?M~OB#NthL<(`Fb$uF)R}o_>@ia30B4SJ-9yh~uUWl_8o`)Dyg~!71V~8Jgcwh7!eFNr9-L4U2JM9%w3FFSMx-!u}rz?~y{8`=3O(UywcaGReYkJt`E&bx7s zZ2$fr;^Vay;D$B2_INd_ye)?QHsmT0`Rr?M_{nr*&qymXxyVgy$wAwUggpDi4Cw6P zLq;d`yy|io3Uxp?;kd~)H~r+i%YvNLywtSN&_5O2<+%mfuKNbG-q3nx@6eq>+oa~~ zi%J3n;fQWOU;(Ew`|(=tPtNO^+`1V&=JaVFcF>IKXeX6>4YxMKpk@12d#*{AtDw$G zwI2v&#S%gcyZV}{osfCruCMBDl5B*;xedEQO-l1EuQv}G3D4?Ns-+1AayeLbyr5T{ zz^Bi@F*(n>AU8-pnrgxmZcV9z)JHCNO9ww0DOH$z((JdY7nE8hl4}2FoO_|+U-&uJ zW(dkRtphbw(oLC)c+=~hT0Hz?U`V?CU&=JVvb`3olb|};ULWUGeg>Z){s|&2&Yf!C zX4=fmcbZB-rrGOL?d7I2sHQp=nB=s$>kkxoR~pj+WW+01pq2c}-1u!p)PdyhOY z_Sf7WnW3@^avntdLD>3oajq#^F3vG6Qvz?&u!z2+&%J=k^Kr)hOq7euO*ZhaeEd6I9MN~eT4>v52rOkYP073EhmG3|bJ9(EsGWx9s>+c6 zCA5lbs#pNRQN-r(>y;G4?s|GEqMvmnbW@Gbp;ryFPFLZYT{Q%CH(gCN3~y7W!Mv^- z0y~*x`%%#AI)m|5!|+yR`EMA)qg9Jb;V9((4xw|W62olOw3eR#Zg}HXQ;yI!NpRU>vMcinl&>>`pd%Ls256 zCz!js%GJh~BvOIfA>E;_R;oi?y;O&~WXho%Fv2F-Pr}OHlZz*L9qF{w*zVM@dZ|@} z@Olr+V;gfX!}7n4Q#sag^;7Jh{`b{acH>b0eDN1!?{-~4nUSrb9v6S6auP=4+pF|u zc(8Tzax^=S3&=OZU7H|F8IAqGCJ4h0EQ(FjL7QwCnrbqlNH>MG!6Hq8-XJ*BO__J? ziY&e>fH&lSb!SVYIs)=pSW|v&ZQm8wm;H}j(HkbL-4&mLG20c=ubN%)7mWFz?TVqG z_}{xiZ)GO`xhq<^1GX#tU=QB4D^k>5k!s4UPVcxSl$WlvEd%+Wx&=0vA-V>ZpsCF; zw}_{TxrZwt%0yWVhE}=Ys~o*-|5iB#=z)TtbM%-k4g;ue!9?a(HbFhkEw8G>38>UQdgIh@4KhpPTHkkwY*|m_&HovB||175YiFUu2GG0XTXNx>p^gpQnYRqeSRy&**_x=eO9$ zp)wc(3$XI0N^EZ@)~kfGomim~745`gm2kHcvsI#@otUB$jqOB|N;F&BnP`=1X(xKA zgwjrQR*BYj;@@okhiC|I{o?gl}1XP8uTq@{KEOOsyb( z`6c~kY+}zp>*vsyblA9kT|b&bNd0e1qEe}6TaOkeP_>dH!QvrB?NQ%%SxU#SjwBYa|X+#&g zaMGi~DOxqDm(YcNK539(bNw+19G24P`}Ya`T<7m^5JR53ADV5O-cWxlqo%+Oj&wP$ zgHvA+{WE1yz(iXo(eV$xH~i5_2yv#fq#lo{fX{T66fsLB!0{|8Y7PyWHm_@Mq;jK2 z)X>~%qlIIx@@WIR2>G;m)_uYPSJ-1;3BvoXYqK-_gqg0iCme!dP9o0s*{};Z+lVUq zt!3}<4@N^DS8%rAbe9U65Jc=j;901ER(0y6A3wQ3c#1~OOBNzrmU+d3@Sy9L`MWy_ zZ_+n1e-J*WYyNhhu-|p$Z~FvcfGhW@$$r8Yt|L#M5rwzutItFWuhQCQoLxQeMF9K< zz^Jp2`}gGNmnfRCv?(AQzBYiyeCd*Z-E)kb(SS*XidDiNTHi zJMeG1-dO&MC_eBCP0D>(bS|Uob6-O-Y{d#OW;xxlVzl`7X8PTVHR9!$=~FBFi8nV< z$I7ANsZI32N~8GUCi>;d$3^!NnrJ;NwkFZQyk0cKwqBH`vTCJmj+pT@tBzc?O59e; z`Y5NgX_eWOPvPVge`IpQX-kL46DIBONLXbAMz;T`%%jw_dJcBKdUcW7{T5%>^?7rB zU60FOAv&{}m!|yao`oAR0p-G@rH$7zmFG6nz=CJRQzbBfa6jeZ2D+`lEbiSv&lVWP zk_|A6DB5Su7{SPigZMS=TSXW3l77Y?s(2Al1^A?+7J#wbX#mW8Ql|bC}qq;|t zfua;*1H0}-f)T&Xc@6nr01J1bN+2fBua}DhnoXdpY3X!SH;6Za`-=YZ07r5-VAZCa zo-U3Ecw`;+Rdcxuw3$=`cXRJ4M83r#bly}!r z+xk$e5ynx4-&I+(rh(mLd))&CmvSbW`hdl@AJlT~#N8 zGyW)4o0l9h+{i=N2(Mu7agx!&+i`iGcuNdU3+xw}H$k@p{8_fFT{u z1ZgbwdYg4*0ttj?MN4N;eDC$Pf^7K#;?jkXk{@AUAe|uB=?Q5B83ykc-9LN1%RpxR z;`O$Gw9ygU$BIrJU9^3vIA2G<-M&$jB53yeW-%~=Zhe1Zh$TXT2|`O41{S?(H@*4( z-$i32U9_X07!yegK}eBw$BuqeG9o3kA%W})#`127SDC=~b>$eJNQq&e9RqNIXCQJG z^jVQK_=CBT(h!MhTG{o^Wja)+L%3pw(icB?B*ZyVV!ASj3>N)N7~%&7AsGo0!xlJ? zSoDi`Q@V3Rh;x+0a29yNqR-t;&+LRLCD92Vt`WA=V;^o1D(I|Tj|wGp&#v8YeRB5( zc!atFS0`QNd_s`u1?T_uUI*V0cpt1I6I{z-TsHdTo((kjz#Q5lg!+4Vfq?$LXDCpC zG^3%DHR~0LHNwGRZ1r!c7&}BRhwGs~hI2VaTh!-sjFU!O#4+pwhPQAGyMSRQ$FLg^ zTZx+!unm|&Gskeq7{10aOyubb& z>))231_ObQ6%EF5VgS?;Z{rE!5_oQZBb}S;pmdE_ZTS z&gDm3)^Yg>muI>Bn#)UEwyM%fu5scPm!0^4x^o%GWhj?@xioTV;?g$}pB&zSkLNh? zF1{~pvbzWfLOXUne`uvnlvmQ<-DZCo-aKJ}(8tz{3bJr$DP=4vqswbX`jdAg6~06F z){OFR=CJu4+6K_XZ%1$si#Kv2~;TegBl$evoF@jqkesAUK6`g{Cxw z&`;`2;wzQ3t!}(nQA=a%$BWWuuI&1ig4GuQ)mpv!5d-l>;%*+&tS=TCg&zqbzMyDm z;zxpsFFJSgA@CGoK?*bGo#F@o7w`q_4oje;9>f>9yZN5P7s9)FNYB1lY7~BurhP$G zX-tAp{2*QX!mBds)erthP)08t?cvvlxM@&B4?4bKN^JWQb5l%g!k-L1k?}ZmTG=o+ zyt&fC8eopZ@*;c$xDT{c(%54Yrv&o7z;tlI1r#8%M!vs%{87gLXF$#XFOmQfSV|8c zGg}qz4R6g@Zoo%?g)QS-+{fPo!$BV3!D4~Pu@cJoaDqQeVF_bbLgG2!zf5Vi_B z?-R4N?4X4i`DUBur_R<`qw>B*ox%7!hr@t-;YBqNfJXl?iy(KabndrXk_R4r?);x3y8~f5wdO(#D*bh;Or2gMc@6$B~3@8fVPm z&SesF_dL52@(n!62a+pZxa~|hJ26~}RMoyOVGqKCJ&r0Pfo?quyLkz1ItyLGQha6> zTZ-tE&mg*M>5HG4rx-glJpv~MC+hoapl_p{10DloOj2Dn61DsRJTA9$1NHyh9Bb*& z-CyA=1ghOHs?D6d*(%`q1Qvb13BI= code[i]); + } #endif srand (0); @@ -120,20 +125,15 @@ int main (int argc, char *argv[]) reset_and_start_timer(); const double t0 = rtc(); - sort_ispc (n, code, order, 0); + sort_ispc (n, code, order, 1); - tISPC2 += (rtc() - t0); // get_elapsed_mcycles(); + tISPC1 += (rtc() - t0); // get_elapsed_mcycles(); if (argc != 3) progressbar (i, m); } - printf("[sort ispc + tasks]:\t[%.3f] million cycles\n", tISPC2); - - for (int i = 0; i < n-1; i++) - { - assert(code[i+1] >= code[i]); - } + printf("[sort ispc]:\t[%.3f] million cycles\n", tISPC1); srand (0); diff --git a/examples_cuda/sort/sort_cu.cpp b/examples_cuda/sort/sort_cu.cpp index e636d525..b8e9edf5 100644 --- a/examples_cuda/sort/sort_cu.cpp +++ b/examples_cuda/sort/sort_cu.cpp @@ -351,7 +351,7 @@ int main (int argc, char *argv[]) createContext(); /*******************/ - int ntask = 13*4*2; + int ntask = 13*4; devicePtr d_code = deviceMalloc(n*sizeof(int)); devicePtr d_order = deviceMalloc(n*sizeof(int)); devicePtr d_pair = deviceMalloc(n*2*sizeof(int));