From owner-lockmeter@oss.sgi.com Fri Feb 4 18:30:01 2000 Received: by oss.sgi.com id ; Fri, 4 Feb 2000 18:29:50 -0800 Received: from medullas.com ([216.23.129.226]:35590 "HELO medullas.com") by oss.sgi.com with SMTP id ; Fri, 4 Feb 2000 18:29:39 -0800 Received: (qmail 3388 invoked by uid 99); 3 Feb 2000 18:29:36 -0000 Date: 3 Feb 2000 18:29:36 -0000 Message-ID: <20000203182936.3386.qmail@medullas.com> To: lockmeter@oss.sgi.com Subject: LinuxApps: Kernel Spinlock Metering for Linux IA32 FROM: matthew@linuxapps.com Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing Hello Information, You are the primary contact for the application listed below. The application has been updated and is now displayed on the LinuxApps front page. Please visit the site and update the application anytime. name: Kernel Spinlock Metering for Linux IA32 desc: This patch allows you to build a kernel that can perform simple metering, or record-keeping, of spinlock usage. Best Regards, Matthew Tebbens Linux Applications - http://www.linuxapps.com From owner-lockmeter@oss.sgi.com Tue Feb 8 19:24:38 2000 Received: by oss.sgi.com id ; Tue, 8 Feb 2000 19:24:19 -0800 Received: from deliverator.sgi.com ([204.94.214.10]:113 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Tue, 8 Feb 2000 19:24:05 -0800 Received: from cthulhu.engr.sgi.com (cthulhu.engr.sgi.com [192.26.80.2]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id TAA02777 for ; Tue, 8 Feb 2000 19:19:36 -0800 (PST) mail_from (hawkes@sgi.com) Received: from sgi.com ([169.238.112.29]) by cthulhu.engr.sgi.com (980427.SGI.8.8.8/970903.SGI.AUTOCF) via ESMTP id TAA21811; Tue, 8 Feb 2000 19:23:43 -0800 (PST) mail_from (hawkes@sgi.com) Message-ID: <38A0DE36.FF96936F@sgi.com> Date: Tue, 08 Feb 2000 19:25:42 -0800 From: John Hawkes X-Mailer: Mozilla 4.51 [en] (X11; I; Linux 2.3.16 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com, torvalds@transmeta.com Subject: [PATCH] spinlock metering (2.3.42) References: <200002090316.TAA00684@pc-hawkes.engr.sgi.com> Content-Type: multipart/mixed; boundary="------------C12AB0FF073698BFC07E20CD" Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing This is a multi-part message in MIME format. --------------C12AB0FF073698BFC07E20CD Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Please consider this SMP i386 patch against 2.3.42 to add a "spinlock metering" enhancement. This patch is also found at: http://oss.sgi.com/projects/lockmeter/download/ This patch also provides tentative support for lockmetering of Alpha kernels, though this work should be considered "in progress". An earlier version of the patch, prior to some restructuring, did compile and execute on Alpha. Spinlock metering is the runtime recording of data about spinlock usage -- how often each spinlock is acquired by each locker and held for how long, how often an acquisition attempt faced contention and had to wait because someone else owned that spinlock, and how much wait-time passed before the lock was released and the contention went away. The new functionality exists in two 10KB gzip'ed pieces of code: a kernel patch containing two new files and several modified files; and a new "lockstat" command, which turns the functional act of metering on or off in a metering-capable kernel, and which retrieves the kernel's metering data and displays it to the user in a human-readable tabular format. The "lockstat" command source code can be found at the above URL. Lockstat uses the running kernel's System.map to translate the kernel's reported virtual addresses into symbolic spinlock and procedure names, whenever possible. Wait-times and hold-times are displayed in microseconds. After the kernel patch is applied to the kernel, a new config variable (in the "Kernel hacking" subsection) controls whether or not metering gets compiled into the kernel. A metering-capable kernel has essentially the same size as a non-metering-capable kernel because the non-metered kernel's inline locking code gets replaced by procedure calls, and the multiple-reader-single-writer locks get significantly smaller. A metering-capable kernel is negligibly slower than a normal kernel when metering is turned off. It is 1-2% slower when default metering is turned on, recording wait-times, and it is as much as 5-8% slower when the optional "hold-time" metering functionality is turned on. Thus, wait-time metering is negligibly invasive to system performance and provides a black-box look at which locks (and their callers) produce the longest contending waits, and hold-time metering provides a more illuminating look at which callers are actually holding these locks and for how long -- more interesting information for the analyst, but at the higher cost of degraded kernel performance. Care has been taken to minimize runtime performance impact of lockmetering. For example, the data structures that record the counts and times are separated per CPU, which means there is no cache coherency overhead when different CPUs update counts for the same spinlock being called by the same caller. -- John Hawkes (hawkes@engr.sgi.com) http://oss.sgi.com --------------C12AB0FF073698BFC07E20CD Content-Type: application/octet-stream; name="lockmeter1.1.2-2.3.42.gz" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="lockmeter1.1.2-2.3.42.gz" H4sICFp6nzgAA2xvY2ttZXRlcjEuMS4yLTIuMy40MgDUPGl3Gseyn0e/ooJjG8QiBrSia8dE wjaJFgfkOOuZMwwNzBXM4Fm03MTvt7+q6p6VAePkvnvu8/GRUHdXdVV17dPD2J5MoH5lhh7M bSd8qLca7cZ+a+/ctcKFcAIzsF1n78x1JvY09ERjJubLzMr6QgTC27h+p16vb49du5mF8FqM ANqgH3eaR53WEbSazeZOtVr90q21154tke1D86Sz3+7sH0pkr15BXdf3D1q1Q6jKD3obXr3a AYBb8eiD6QkYK9RiDLYD2Y38R9/72AgeggZOOM8D8M1H+BlCZy58n7A8uiF4wpzPH+HWce/h fmYGEMxsH2amdYu4hd/YgZ3q98JzxBzmLg4yR7Yz3ameXV+97r8xLq7Pvr/s3fQGO1WAIe8Q uCAcczQXSOcqZA33sa0ZmOOxD+6d8GbCHBPM8PIdr/RrhGoUBoCkufc+04nzvhBwZ3q2G/rg E49+YFs+hD4iRbKFBMYJsNzFwnTGjZ0qYepPGMOYZZBi1MUhmAlP1FgyV7Qc+sPzK/DDEQov EIsdUFzSMImMp8ulPkp86pkk9qHw7mxL+HBuT+3AnMOVCO5d79Yv1cBCBnDJ4Ko/pON57ZmO JSo74zUqfWneiomNYivQoWhuVVWjmVW1bDc3qmUMmFPB406rlajgSat2DFX8qR+w8p29e/f6 ovtmCNUXUD83DDw1w9jBA0euUFnqZ3K28wK+LkdrK1D/gKLAn37g2VZQX3pu4AaPSxRb/boF 9Ym7sIP6xDMXor50bZSuhwpWjGm6GdkOdIvgSBXsyVhMYFVv3w97xutB97L37mYAL0DfqUpu EMAhiPR8TBaxX0h2JAl4gqop4Hl94rh1Rao5t01S1+c11m/XQdOzJ6y8qLNLPAwPlcaBAI8G 7AClHW/2ddlHfzGn5cjWGQqiAC/Uh1B3YW8s7vackIT0YKX+epl8bL18pp/Sxg4Ia+ZCqQhd 6RQm9np9NT1rtmfOlzNzz2Kn1kAlL1C0onWrely06ot1uhDJB7TB71CqrUPQ9zt6q9M+SPS7 dXJY24cq/TpiBZ/YdHgj153D80tzalswfPQHH8nnPo+U57L7pn9mDH8eDn4gLYFfofS1mkJ7 KKESlR5L8LsUMHszRqcc6QW5w0vlDp8XKCSRgGqErjzcLHy7fXy4heyzy9aIPruoQPLtz0s+ hyPrWA6anVY7JfgDKfiDSPBoMVJM52IUTuF2Qe7f2rudeELEUjrvffue5H+B4vq/PaVM0Pr7 pyRDIX82bv3Hhd+wNouxAGDTyRUs/xtnWIQtd5qtjn6cylT29Vob8xT8pbf4OJ/EjrD307vr wQ0exOW31xflqQiMe2tmOpVTcspP8GDQydqOGJeTw6nAs2f54Vj06MyzKA1/aTsGHZhBSIsm MevZNB14j+vmMUMar8V979mB+MxsGrcSyjp1mfh7GM0s/mEsbN8qVpKCZauqUbAorxAY6Te7 0yIcaW/aJm/aOknU4OBEp3SVfu03WQ00TwSh52B8C53gdAc+UUb5ZF0oFg+4rQO+b/8LBQek KyQ7psawnYlbRs3xYBdTNl5RQzOdTHDlLkk3B70MPwMtNXBvF21/F24o8Z2EjkXJM5gW5nQ+ phQZNwCExFtwft0AAtvbqXIuasW7xvrCQGUMqaEVAOd1u/QT00ImYhROMNHVNA34n4JmMSVc LZeuj+r+By5UciwQCWGKABkA2fqU4uwDqiHnyJgZ8HHG8EisLwJfJSAhJTB+MVeJohewBYov 1/EDxR3k2Nuav4JDS/iLOYuoS6gw3KXw+GB8YHVNkKRmXgBtdfX+4qIGyb+9XZhjdXHLjGvZ AyQmcszj0KdTSQVvbzvueNP+K/OSimdryVQbSGfB5OUthSmlemULQeCP/DgRABvFAFIMCrbG xn2os3Ef7h/U9H1p3EOj/3rQewN/An4avB8OaqDXoEn/cRZ/SyZvLdcTK3JAZ3C6wRdkWWM8 Y9szsL71HuWfnuumlEWJFTc9qUEpHi7RCebofHMtP33IUlxNKF5/evJslKvBnL/AIg1S9vKd a49TPh+25EeJXJ0REnSE3KjBUm1V5mt44QM70inFOtLbWMPxeWlyHzHF4hmt+Fm8bcR3mgQk PoZIDTeIOySvrFdpxJgLp4IGT6PupBw6vj11uB8RMIJPG454O3Ji4ZI0tYLxiKJ1J5E+hU9r g6/tWPNwLPZMf1GXJUSywawoOm4EWA3IG5dzWD0XFrR10A87zSb+B/3k8GRdaP48tiRINw86 7VRXqVlrYqZWO2CdiIIEnLnLR8+ezgIoY3Wpn5ycwNCe2+jU4Y1nLme25deg71gNWi5BKLAE WEGOHuE7d+bAW/P+FkNmeca/X/lTu4FlbUUu/tb0kSjEdjsnS0CREhg1moaBwDzPAwwq/CED uKeSRK7Eje7Fu7fdRIOMtzgn08TCOYKUcoJ/oKD2ZvfectSYvUygzt69N85+PrugCr/3w/ve 1dnPGq+qv7QeLfSaWH18zCLiw9i7w1BJWQAjQy990b96/5PxY28w7F9fGWfX5z34B3zfG1z1 LqLRcqvWrjUryeZ4YubcsL2Phm/eifJDBX5DBTf4L8N0xoY1t3G0CAADd4A+NYFRA8Zkbk59 CcQar6FXvyCKQVEMrUar8aAEq9yYgcaCbAnDwC0wU3BcxxNWiOuREA4E6KPYtMrePSezGLvp t4rdDBSIxbKGZITzgKMjEoUyR5SGcefO0fehNI0yjpe0+XhuzOGpXnva/M0p8dDCvYOv2zjS UiMqKI3mI59W6pPsuB+OGEMrgdD8YG5Y8LSVoNU72mIkP3eg9GJRgrJhjMPFQnHDPFTQwb54 5uEcsZD8JXmpSNg1oDjJXkllLxFMlKdsI9+4SvgPC1gboUalJau5Hq/Q/ytEul6oKi9CxV61 eanX27v45TSgTvW2Dj5evo17jxcX1F37G+uuTbiy3aym3tlvZbtZR+luVuQ5qKA3zPHYM+7M uT0u08eKVtYr3HdRi2wXjX1hLo2lOUW7N52pQJ9segG1K30CkZVTjdKVgDxPXVsHcGd7gRG4 xnL26OMp2y4vLCs8pCApTHzWfwkP5jo5VCl+lmhXvcHgelAWRCxg7oBZyW259NTvPB13YGSO aQ08beqH84cGKnUNbel1H6OBQZ/Qp/MnwoNiQyzrm6GrR0aNBdLobXUrWb+NciWr/w3alUK2 Ub0O6FnAgUom8xFR9owoIEI26qKj8zDL5ZmdenQ4spM3fNe/IvMFPT8z+KDGq2sgmvkZBdEk FcCsBjDrGNqLJebSxJ+s5JM8vgFY8gtP8HO04N7lZ0umE2CW4yJKjP2m50N/8MNzX/J+zJ2N Y2VYcWFbpk0r9ZeM/hsoUe/eKQH6N2pcjlGlohWuY1jLsHJKqviJnoIhZWXMB+CJwLoryXJW 8uSodJBBBKiwgFRXqxwdHwUNkFGDi4fFiFJfIv74hNzCSbO2r0e9ViW6dPtLOmAofxVgMsEJ iI8Z9cgOys2acs6VHCRv6947zGMN8Gf3BlMddC2y/kH26om/zp2gSu6I+7VFroqgzHOqz8fY ZYg0lgFXBk8gT5jiJw2mYmqBQDMxOUGR7FOh8jKqlWVAXixPkyHk4D0ygolJ3ReWLLQDlxoa 3F8xrSA0qa3sLsGUQyiXBBz/uRP5+NUd/VOoav454kARgMIIvgumRCu8CY5l4Ef0gHGGflCM bUv2qFiG0fzafAFjuswYAH1h7akeZwiUfMnBVjpLADWYzi9kppBHID7CCoIofSg1Iq4axGOr 9lvJfPitpCZbTFMEXUzTKB5UNOkxahTCHT0qjvIUzkVUKrI+8eCspGgun9Gl9FG1los0Mq2Q ylbTKhlBfolSFpj8inJKu0/pqnWr/NMLaEpGVhNTamVkm+GKoxVOst4iB6N4SWz+q0KzlvOF LoG991//Rwj4iUSzRT5Pbx/F8WqSCRQ7kD7LpMeXysClMz3Nroz7gwUL65Erj3xaJjIx41/i 5DcQJT08uyFPTB9O1VOYfb4tctyMHtdvyrOB8myQKTVsdMI5tU8eiaTI8u7TipIWUiUHUcmh Sz1EWYsvI4lKHqaSLQsKVQ7yYSg5FNggfGVvK+LfJg/kR2lf0ETKr9+cB+ZX/60W0gqy1NM+ vdnBvPLgeKWDhDni5g5SjSD+C9tIfXquWdxFWp2KtTVubbLVYSpnzP51urmVJBf9R/tHawN9 aRn6s8kcTmHpLucYP/GTNbdLndKLKXqIh0qHbMNx0QgoZ9ndg05pIRau91iS9vWZJtTmjdWG uPVkXop2csMg3krRkNpymwZWHJjSLTPugVU0HpP9MDnC/JZzzmLlxLdvIbDR+FhhbFXlZVdv Ydvx2oIK73jrCi+HKmfXR5391GWY/SaFTP7JFw3JN65mCuwd2R3Kkk9m+hSQ6qTJulRoHf78 c8MjfNBwTfmralVVhzJBMcZiGcxoVktS8GdyGz5Kilgyxqqa7ChVk/2hsBagRGNqMlr0/GUO 1F9Iab2+CWs6NSsmd3t92rptkFu+jUZtahp8qUoluFZ0qr2f6NQh1834s9XKdKTSGTDdO0OX Si0afnCHjmIUeHP4GhP/ZmmbGwP5+pAK89WEfG3yvmEx5cXrE+PEA62pH2WSjO6x0DVy8z6h OpYDViJry5TK5l0zdcbn982fwGd2TvxmcaIIsdRyrYriqoVtlmwwKgDkBS9WnKPj2glUT/aj a8+x7dIzxeh60YZniytl/6nKyxFBcSWc9jpKIjS2QSCZB4uwnvt1VZu8Yst3pqq6jhmVujTD vjGp29Kea2vOM9Xl1rxn9OHLuK9v6in9ARu6SqfwqbKBofV4tXXVp6zC/mqPK5V6FNcScYNx gIVNnSsFiA4X81u+uE72tAjngU0dSCqA6HINV2ntQ9Lpqn64X2u1k0iLu1FXKMID9H4AwnCD 0nZg5XaeT53wGSZgjaEkqrqtn0zKMe5g5ouuddVZ0dpIGwt1v6hUpmptjdVLvffu6y8XfKXz qxcgizR5lfNLrSBbdGYsYBTa87GRJq8GJSMFMTHRLsalbQy8sEgnLvmsD3R51scntfZhYtz/ NiZzhXABlxn6mM0UzDo+M3V8XANXNA19B0R+oxwFa0oOML8vQZGzeEaAyo+l+8e5+rpSjNoc jzEPKKGMmG/j2353aAxvBlCi1GCb/T5vFOo0KdIXPvnOna1yMvRc1gzchW3RIN9Po1sxyVCF Fp4mq8bCKsfX2GSiSn/VX6qrd/CSssmqFj331FOwKOEUrFqgenmxCW7BKOUy2Wupn2m5JOxX om1XEVRSVfI6YebR5aSptHYVyWfhpNZS7r/NaUjzU8N+OOIQQLGgnNaw6L7h5/s8sprfssmz snh9zr6y9C+3d1YxbdPb0Q9P/p82d2Qnpbi7UzAXs/ijOQ+RsonrcRgumfxsogRLk17wIftc mj7RF7jA5BnhcmwGgpnVfvnlF6jTG2X3aDkmKp8JqLN1P+SLxNxfDH1+3eybb76Jrg5HdF0M b7o3Rvfsxri6Nj50+zea1iyapXipaZpeOHfRe4dwWqtoctD7rneGs+2iycvuT8aP3Yv3vaGm 7acEMlwKyzbncJcVDKY20ALMmHx6YGU6MOgmkmHAFemssDroGoMP67jByWHvsit5iak5o9vF 9HyWXqsaMzUcIugagLyqjTkS08e7j210PYHrPTJwQF2tsW2ZnKdl2HIn8cuDEa5G7nw0Sdfl +4ubvvQR3fPzQW845Cu0KRovzQd7ES74PkIaMVHkA9/gQORoIBhZfCEf9jEZvCG920VSddwA xTsF4bjhdNaA8pWLKzGbpPsKM3O5FI6fWsugyXpMPoMZf/Ddhdoe7u35XOIVsHAdm9p24+Kj IW3gD/2r895PeAbcFCAWJZFD4g2dJixM/zZWipnpzySD5LRdHouPoHijt93hW+Om++1Fzxj2 f+mh7jVPDjXKuRehz5S2dnevNgNedoffa1q5EGFd57AUHY7GF/wlIp/UVl68jZRGnVKKaOjT dXUnMG2HL8Vrqcv/YI5cekFVqs1IcK6fkavWTVCprfh1vwhjonT2GKftCSknYa6zmDWOPz75 wtA3pyKrph5d6FkG/NDmni8xUH1Az4Tr1jJkeLQ9U901DumCgwsL27EXdHiWac34JVMBWGpM 8X90QPQWJflQdUeZMhxKcbVdfpPVMzzzlA8ILR7piV+WCOh9WiyjJHV8iV6CRe0UCZU2Mbkq 9LF8CTQHo77BVotGKh5ONfkoXXpa27lNyiipzbzfzLwTsl5C75zWP2lQhP9T3hfIy9ZGkHrv A8+JsdfH9gLtCtnBfU3PMx+ln0ES1JGBOpLE86h3RBpKvYSCQzVrJ+jMeYfntTo9IQAnXIww CMP3Qixh4rkL5MFD4kmDkoPxUaeCe4FxE2H8CP48p1AsLdRTpUAiUaoIoquuAfSV8/Plu8+3 CEknSJwRLJ6ha3F/k6Ov6Uh4UCyCMmhu90WhnRlmClQYyNhTovu2n5ySKT+3nLGEZDxKszPg StQszw3KGaKvOdw3As0KF8a9aQdGYKOCSHXDsZCKiTu6k7EgImlFpHlpyJmLZdJGSFqRgmy3 EHJhPqzsiWO8C79ugxpQAJLfjEBobA0IS+LXomD9+2mi4WxoUmhKuyNpFcxrcoyeTUVDvxb5 /t/Tr0ehPdLrQh8xZNK781Kn68AFQfw2Pp4b+gnpFFxnz51MakABwpUvGhWGgevXr1eTneur 1bTg+opTo2HhzNvri/Phau6DM92LiyhARBjgzxxgpqUqp9QzN9qsUPWk59KULNWjKXmi0XMq pdjsiflQszC2b8ivSRizw2O1l48YWKCCLEdWXGiVDI86gdCoMHh0Snuu2J2wLeMYXf3xCUFq OTtXuorGb5skDjY2XdqYFgEtWglajIn00iBkikCDBiJbkSSz6qL1+sLCbBdrD8HFSaLNBE03 6tE9K0j6y6egRUCuo5jMPGXVNCUi459Yh6GHY9LVZ/Rlwkmr3D36F7W+EFdEfRpXRL+MGiiO CHkSxjRtVz3S4Y4NX6Nl2CiWIVR6QSEolqkJoPK7KUiQkGX+GgF2uCWEKFUkrlWDxZCc/MUR FtGq7yCI3HVIJjsKJxP8RUFLKhMHGuleU7hXQqS2mwzFEXzrDRCW35XMhGFabigXorxU3LIw DPWs21h9ayNVur5cNdRBt+xUNHpwIa8c7lbKUC5nTr4iW2HolQ1JoqFOrtyswDP4n3YFPYJT gYwbOO8PKKEseyZdZiznUdI4QiB4YVJaqcrsM365VKVusjEh6wFK4yhvQZTJl5jITKsuM6wo 6AqvFmf6fJcwUbwkjQx8MZ9AGW3Q9cbykBzbEphz+kvqZNHT9jhXl/UGZQXCrzSwTJffQBGi 2ylRNGJ7LqW/Z0XdWWQ7V/vK3l2NcVFFgMU6f6PEBj7IqZH86qgvU5c6A41IQJ7AMVGjAnqB 6RanXJjC0tcARJkt5kZ+UMNwuRSTkL63hlJcE0nATRiNsmacmYn5WH2djPrCGcTrC2qiEyuS dMkNfRMMqdeCjp12Q3KVfJKE/56cZJQPrghKMtGf0DLOzLm0S8MTM5gjPSxRUGKcNPnlsdWk 4GgTdcxcLPxve0fa28Zx/bz6FWOhUEmLlCWnRxqKLhxXCdwmcmorhQPDEJbkylyY1DK7SxGM 4//eecfcs8ulYjS99CGhyZk315t3vzd1iVIvyk+WSqk796r0TiGO3Gk4A3VGfSoxsqGVpIsK Q0Yl8k/4eKZzDuXI6y0AzuQ3wtca5qnRgUSKoNTgA0kNUHjEAkJS6aQMaYhLhUUDA5hkUkjP i7JZ0Xz67O/fP395cX1xefXy+cUr8dljJRItstt39VytmIJCxCGe9iFbfJq0FV/vsG6Lp6BQ S0mVLP3mEeHyZl5I6ffHdQ4ycV57Uhn/ck2MkYQ53VEepOwAzEgDsMmg6kuGzFHjSsgQLpB1 GiauehMj1/TbBZrY/3zTuN22BGng3hSOgiS+Jd1HsoViYeGGUoPQUKfVNVcerFgK1BQC1FB1 ty0VQY7ZIuWT8IAy0Yjd1rzMhKeF66TDVkqb5DUSM+HSKJ4UzIqFMgucYXltACU+AbzbbIMy EupdnsWjiwBmZK4GzemTi0NtrF4uFmD5oyuR9SEwDalf055Y2jaggcQcqBcmeX41TW+rFpkF P725fHkt1VqJfgDrKerM8m6ycKGi25HeE6O04Llomjy0/23giia4+g9B0p9sq8djcEOkEzS4 cC6vQjh1QwRU0ZJaoiKWhh2jPRKNpJtb2+TCQcCA469fv76+fHH1w8WV/GpVpu+WKZpR5GlK xpLWErcmwGi8oYWLppL2SgwZnz3+3HEFaQ+2Ea10woWOfQvs4dCkm8sD87Bvqi4OD910l7tD NwzDk37XXqmkCU5ruNvZ2edUXO/zwRkFXqgKIm2FHYplXoZlAUbdOq9W02u4R7H+Heqi3K/g hOOm58AF7FLX25MpRR80HTrHGrTVpvOahIfsNYgcbntdIr+/m93+ewnCym5/jEUq5H/5SF9c v/jyr1hCbZ3Pzv5wUphidcf5Tfaj6P0mDPobbKX0bnoaDaTQxeEOhNf7b9+++At2tIZ8vyxm ZsQde2xGiVb9CVs17rTVZm+3YQMIt5rAH8NqAn96/Ov7C6d7+QvdgGysIMjh2O4vWVneFtFf qERZ9KdKEvxZ/JflKvp9PQctLT4FN1rcg6cDMZ/4mjNVsYz8sCbDrlKoPYhNCvfTV68uXl71 wFZjRaFPfsrKAkKnBqCj9JNECuuSi+E3p/Qd8crz4f5/cFbAoO0/+540/1HPJ/ccU2djebw3 cf4NRVq0FXMgPpyaiCYqPvf88vnVRyoWxCQcyyNdZ5DwNrKGYUGLoC8loTO1YTjadCDsqFN9 0ZR9t3i/XuEX+J+v87sMtOeXTweo5NWZJ1uSyGlcumzYf+RP6ODYHqAXeHYGoYuIQ2O06Ohp KnHxc0UZ/2SUHLu7fsKy5xtlkjFDgQYDUvsq6AMiLZRuoixbaPIGgb890b0h9MqAQlUDg0N4 EmNxyl9y4Esvejp6JyxYGKQDlW8SvSJ7Br6/S7b7SMu/caeqYMNMwdUMVfFwQk2N1Oe+oJlH 2xEsMyqvDpvpPEcPwdxFk0OeqttKqVoqRBLXQB8io4rxFoU+4Dh6eXsaopmQi25EtV2I5eHi QMz1vnvqFeaJECoqhyUaVFDV2eTgWIcKbxzTtNjKz1B9uSQjjJT5daW4qRTBSMuYFeJDLD8F hkdK1RA5fhQgtCX0Q8gspZD3TrlaxlyhWnhNRjuvFnXuep0IFpRo3H2vul8CvgDm+sEI7jQi urQ4jwYuqPscXXUEzPEx3tlg1eHqrMW1d7EunDZItXbwtqTLYSWtLSQEj8IgMjYmDexAOYWw bq0hxmYHHeM0he8+XGJ18SlISHRiLAM2qVBglstm9DKhgdTkR7qtZy5zk/jUfe9gXWAOxTIQ zUHjnuuZ7RtkdjeU/HoStbXQQGhKmwo7ZvECBwWY/8aYDXQB28VYSMkSirSBZFdItJ5hFnjS e+hOgo0wssvbvsZu9DLTot7iVeD5+24Pcw2OxGd9uqXwd3TEy4iu+Mj3t9odeCsfjIN4tb7Z HUmNv5LCym2xGZBpXCJyZYp7sKEfIpxKE5FP0SJIhAGGc5rB5XIsSbA5tAnQES7Nws21dJFe MNZDY9w1B+3kUM7QwycN5lyklSS9njcbyXFToOUjdpv8thJlUSyVARFt5CjapXI/ZAPbk5t0 mwlTQ6+xY1V22jcTuz1AhEZ1rulHnmLC5n3nZBv3fdING/0xesYqszV+zLbYtJO24XLuQ+AG lsNB4Hb8N5G85N+M3nXr5oQZgYGnpwKJ+uQFUmSzAzA3fkjuqzpjIov7QxgLNYd9iLeUsToS 7f9T4/9wavy/RI9tQpzf5jX57ihVXRAdNbme0MAXfm3XYt+G2laGigDTl6APXq+mYmwCVU4t MhlLQNRwzFWzWYlFDhn4IEwdoB2hei4KiGYjWFSPzy9ymNCW/8cb3DpJcycDhtc+VbCS2dXB 7XmIoTNLdcC4/Y01irwDgHYSHSDUrOtx9Kg95Ay1n8ynPpvuQFgo9jHerNQu+g6vjKBmlYMH tKryySLrq8QWvJV2eMqh6DkBD3YcsdBhDX2iY5RBcZPLds+++36AP/NU6nnmRs0gFq2K1Qrf jAJgqglRRs8sJA9OXVpDSEASAEdyz8g6RngKxakgpsMRpB7aVM2gi2WmdE1CWlzCRhpJTpsk KPHwXowNNra3H38ano04iPMJzIc+D4cGPTHRrYU6B7zCth+ijBZsZTi5JohOUA3Csje+G5xI lI3mtYtiUkAS30+ZFT0eZarta8aTZHmUjvZMD4ICAIZ1FesSAmwh5mkFUezL9Hargz8glMSw evnrBuKrILBNdQdJSIPASK9FRrECxQa8VCDOSoaX12KRgl9D4GsU3H+TbhmiSJGUzTnoq8AH plItcvGAuAeYXQpCckxK5tPdS5gijNoXQVVHx274hOVOxQD2xFHKkU0+KIQfkcwQCpWNE1Nm sWMzwX3EDTLsT6Ro+96WWn4lmtyd/t6T9LaprvejtaSsHqCN/au8hIQCeXlhNHpjxdymRxDw qd1UKlYbEJsCrLS6GeEWId4PhL0CJX0ESxDOtJVCZidlI2ngsF4eAmKiRDXPb2piiyoIXntC nHjAXlESBFga5ulBRqAKM73DVIsphPHOuHaowY7+CfX8Rq4Ucoc4529B5AQ1DMgzsm+CplQw Wl4TgEeK65AuLRkIfDhHcnEr9VzMMYDvjo/VFiii8iBKVDAAcgEK35YDbXH2eWXOq/3AkDoa CajDufgHc2ClxmOcfQpvPt5AthOHcNo1PK1qLFbJppEj5TcU9zRyfiyG9KAhUtR2RTnWG2W+ 0b04SQi7KAF1X8Nq9C6b7T0K9/fI3WDWbFnhv7LI1KFDZIAzTSAScK1yOGxe0WDP8UfX+keb aQd+11vj6S9Sb/D3uqsdxc0Bc8069kkYi0JH44wF9NxMHDH3HiDGwpkM8ZvWAlCqlJWluzbU 7uyqwXrp230bax29zSDuzpIarsJLtBUjOCArAdM8Z9m0zPidW01QMRJ/jZn24HqmCjsD0xnY IvbGBwnLJfZNoSoF1ZQokWTnlLWKWbuJDhwJNTNV88vMeaeiZ3l1MLS24AyART7Na0W26AFa jIvGNz9FMcG6xTSNj8QkKbFAYrtqzskCcv2LGUqXGIK72Ip3ZUoB/pjxocPdwdrRovrDGPg7 7aiSFcC2aHISaSqXxYbKqUO7hanwAwGvXOzgZr2g2bPU9KAbBhDGhJaEjjtv2xR22BOs69BU e9a7D00ovtdFid2HaHUT/0LAPQjEQFGQnHSjBahPhbi/NuoaczFH6H86zOXsFMp5BAa2QGQe IjKTnXqYV8NOiNx8dP9iTEarWEtRn/sZx5oR2JjKOu5Hg8VsD1LaZjPbAaab1cxx34E5WIf8 Vat0ynZj2j+JSq8yfJzcZMVAGyfRAB9ZTinPHZ8Hx4zO5areHurcEKH1oGgQR2LitIKwDOs3 JcOPSYI/5yQIJbqTqt8oa1iAkh2GqsSJGwt3rcwWUNWkbeNIc9GuI0qSq7gwBoWTbVQCIdE1 DLAlMRNyZZfZkmDQjkP1sgoK9NRgJGFlCLIZCQhWicByKViSDxvm6YIgGMiUiYmvdQzAfMMv ctsN5umMy6e8W9+yAva0qtZLTvtA6Or4UeiYQPEG43jAODkIlB1Q5/wkOwEP2S18Z7rCRPCB HlPWxmhrDspQhcgY3vCR47vJvXiLyHHzT0FwYGQciX8tY8CvEfiQHeTDDtTPOPL2diBmv83Q g4baiF/alrc5yjO2GrsZQmq+StbyOPQzGDScesvQYePYBIJW/jRs30nrU5Waur1+/dqxAsgr Cmo04eU2qx+od1Ip2Y/efYyleKOicyx6DrCHbg87QayvOsQC+vyOQbineaNj5/PBlLOungGG fFN4xdI8S7tIq1o7gT8YXu6sTm7AjyMUTzApBzi83KIkqe/wW4Kt8v9K8CTxG8j2j/QqjPWz QlmumIDXnd1y1rSwFRpc4LGiYTVnV7sSd+TUTpzKFMnYrWsxchqZUhSJ76+oVNoltObSE0ky ds1E/HMkrDKAF8uGPIC4xGt5WrCHxc0s3faO6ru+AmsXnZDw6ruT+u66yqb6dy4rAdOKFOFX zbzUyWBmfmqlN7rp5jUIakPIJkrSO9KR/nZTVQsi3tDDFzCGtFwxSywLcelYeMD64onCds+u GXQmwNNitYXn12DUnro2etJyGoNgCFhDZCpjv2EjH2OaevoWTYrAdi0++mej5rTPfMc2OnTH lqN22UId99o99jxpnDa6tTIQHrK0XFj6XOQQ5DABgIFo5E2sTsSOitwmnY5LmZyCnb0fvR7x 8Yv74u7PP4tQNBE7ETu2z3FU77LLckjcXbQd/bK70ILQjhvJetAdBFVMdzVZUQKyPznxu4Ly S7O8os8qOYGqVYgJZs+tV44hYbKtdZ2Jww0l4x1iXQjudcXFnL5QcFmnL7Mhjan9SsxdIy+6 2y/Fe7yYTld5irKyLEqbJcrdpsceCbnxI4iu8Em1YWM6gE9gcPhgiI0cQJyzF9FCk4vnl/94 +s2owaYai473A4RkRxA4EH3UsAMuj4OIUG3yejoX+kdlYJ9CLUgdHvmFRV5iJv8HYShlNOLS tEJREK04IMZNyEeERU9YxeEKI6q0gxWkZ+nd+iwuvvz+1Q9cWYPdPXIQqw//XzlIvRVyDa4v gq/RgbFr9ZFA0vYwrkgOQ7JTIsEoAFjgKainwBkKKUfdLIpNpaiFwsRfRP0Shbr4nLqkNtFC lvbr6ug9or76BuxibUoXaVUXyWUT2D36IzRAvlunZSpV1MzVdKXmi/qr2hNjmwnVyveULNtT Sx6Ir7/6jksTWBEGzfNEFqaR8PLFq++ejZqGJW1TDamJRscRQYVtHy3UXBtFhXaNUU3RImf+ JFtEJNQKD1QcXXS2HdRFNQcHidyopn63aYUqa/PkEk0sWKxAgNjMQkmJey9uM+BiywKqntJU 0fCDlGo45BIq8n5uqUoueZqJtMcsUhzDYVE2pFaUxRzDPMNhDNbEmoMkIBjX+vdFlhhgW5Az jI+PIdYhOAc72LLtmPv2wcSvlqcmCVsfivdQDEnzYmO0z9mYYB33IfKBQyC8qaE7M1Gt4TGM rAbh5BBZyKF11Lv5hSn1GFwKavrz2GNR2oSyDye8bGWEs+wmXS9qxeU0ICV8SLSEvSFBgTem a7aeL48ocRIHQRHynzZPanRMlQAA --------------C12AB0FF073698BFC07E20CD-- From owner-lockmeter@oss.sgi.com Tue Feb 8 20:08:19 2000 Received: by oss.sgi.com id ; Tue, 8 Feb 2000 20:08:09 -0800 Received: from neon-gw.transmeta.com ([209.10.217.66]:50699 "EHLO neon-gw.transmeta.com") by oss.sgi.com with ESMTP id ; Tue, 8 Feb 2000 20:07:48 -0800 Received: (from root@localhost) by neon-gw.transmeta.com (8.9.3/8.9.3) id UAA31832; Tue, 8 Feb 2000 20:07:39 -0800 Received: from mailhost.transmeta.com(10.1.1.15) by neon-gw.transmeta.com via smap (V2.1) id xma031830; Tue, 8 Feb 00 20:07:31 -0800 Received: from penguin.transmeta.com (root@penguin.transmeta.com [10.1.2.202]) by deepthought.transmeta.com (8.9.3/8.9.3) with ESMTP id UAA19881; Tue, 8 Feb 2000 20:07:35 -0800 (PST) Received: from localhost (torvalds@localhost) by penguin.transmeta.com (8.9.3/8.7.3) with ESMTP id TAA00979; Tue, 8 Feb 2000 19:40:12 -0800 X-Authentication-Warning: penguin.transmeta.com: torvalds owned process doing -bs Date: Tue, 8 Feb 2000 19:40:10 -0800 (PST) From: Linus Torvalds To: John Hawkes cc: linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com Subject: Re: [PATCH] spinlock metering (2.3.42) In-Reply-To: <38A0DE36.FF96936F@sgi.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing On Tue, 8 Feb 2000, John Hawkes wrote: > > Please consider this SMP i386 patch against 2.3.42 to add a "spinlock > metering" enhancement. This patch is also found at: > http://oss.sgi.com/projects/lockmeter/download/ Mind showing us an example of real-world usage and results? I'm loathe to apply this along with all the other changes I have pending, but I'd like to know how it looks and what the results are under some real usage, just as an example..? Linus From owner-lockmeter@oss.sgi.com Wed Feb 9 05:22:34 2000 Received: by oss.sgi.com id ; Wed, 9 Feb 2000 05:22:25 -0800 Received: from zmamail01.zma.compaq.com ([161.114.64.101]:42247 "HELO zmamail01.zma.compaq.com") by oss.sgi.com with SMTP id ; Wed, 9 Feb 2000 05:22:07 -0800 Received: by zmamail01.zma.compaq.com (Postfix, from userid 12345) id 52103442; Wed, 9 Feb 2000 08:21:56 -0500 (EST) Received: from flume.zk3.dec.com (broflume.zk3.dec.com [16.141.0.42]) by zmamail01.zma.compaq.com (Postfix) with ESMTP id 30E3A1A6; Wed, 9 Feb 2000 08:21:56 -0500 (EST) Received: from zk3.dec.com by flume.zk3.dec.com (8.8.8/1.1.8.2/16Jan95-0946AM) id IAA0000020169; Wed, 9 Feb 2000 08:21:55 -0500 (EST) Message-ID: <38A1696E.8D8E55CC@zk3.dec.com> Date: Wed, 09 Feb 2000 08:19:42 -0500 From: Peter Rival Organization: Tru64 QMG Performance Engineering X-Mailer: Mozilla 4.61 [en] (X11; I; Linux 2.2.12-20smp i586) X-Accept-Language: en MIME-Version: 1.0 To: Linus Torvalds Cc: John Hawkes , linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com Subject: Re: [PATCH] spinlock metering (2.3.42) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing Linus Torvalds wrote: > On Tue, 8 Feb 2000, John Hawkes wrote: > > > > Please consider this SMP i386 patch against 2.3.42 to add a "spinlock > > metering" enhancement. This patch is also found at: > > http://oss.sgi.com/projects/lockmeter/download/ > > Mind showing us an example of real-world usage and results? I'm loathe to > apply this along with all the other changes I have pending, but I'd like > to know how it looks and what the results are under some real usage, just > as an example..? > Not to ask a stupid question, but what type of "real-world usage" would you like to see? I did most of the Alpha work (yes John I still owe you an update - in your inbox by end of the day) so I have scads of example outputs under various loads, but almost all of them are benchmark-driven. If there is anything of that ilk that would be remotely interesting to you, let me know and I'll send you what I have (or generate a new one). - Pete From owner-lockmeter@oss.sgi.com Wed Feb 9 11:30:56 2000 Received: by oss.sgi.com id ; Wed, 9 Feb 2000 11:30:36 -0800 Received: from neon-gw.transmeta.com ([209.10.217.66]:54795 "EHLO neon-gw.transmeta.com") by oss.sgi.com with ESMTP id ; Wed, 9 Feb 2000 11:30:26 -0800 Received: (from root@localhost) by neon-gw.transmeta.com (8.9.3/8.9.3) id LAA13457; Wed, 9 Feb 2000 11:30:20 -0800 Received: from mailhost.transmeta.com(10.1.1.15) by neon-gw.transmeta.com via smap (V2.1) id xma013444; Wed, 9 Feb 00 11:29:55 -0800 Received: from penguin.transmeta.com (root@penguin.transmeta.com [10.1.2.202]) by deepthought.transmeta.com (8.9.3/8.9.3) with ESMTP id LAA27641; Wed, 9 Feb 2000 11:29:55 -0800 (PST) Received: from localhost (torvalds@localhost) by penguin.transmeta.com (8.9.3/8.7.3) with ESMTP id LAA01377; Wed, 9 Feb 2000 11:29:55 -0800 X-Authentication-Warning: penguin.transmeta.com: torvalds owned process doing -bs Date: Wed, 9 Feb 2000 11:29:55 -0800 (PST) From: Linus Torvalds To: Peter Rival cc: John Hawkes , linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com Subject: Re: [PATCH] spinlock metering (2.3.42) In-Reply-To: <38A1696E.8D8E55CC@zk3.dec.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing On Wed, 9 Feb 2000, Peter Rival wrote: > > Not to ask a stupid question, but what type of "real-world usage" would you > like to see? I did most of the Alpha work (yes John I still owe you an > update - in your inbox by end of the day) so I have scads of example outputs > under various loads, but almost all of them are benchmark-driven. If there > is anything of that ilk that would be remotely interesting to you, let me > know and I'll send you what I have (or generate a new one). Benchmark-driven is fine, if the benchmark is meaningful. Specweb or similar would be fascinating, as would lmbench or the SAP or Oracle load benchmarks. I'm just interested in an idea of what kind of information you've gathered, not in any absolute numbers per se. Linus From owner-lockmeter@oss.sgi.com Wed Feb 9 12:02:47 2000 Received: by oss.sgi.com id ; Wed, 9 Feb 2000 12:02:37 -0800 Received: from deliverator.sgi.com ([204.94.214.10]:21820 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Wed, 9 Feb 2000 12:02:08 -0800 Received: from cthulhu.engr.sgi.com (cthulhu.engr.sgi.com [192.26.80.2]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id LAA02669 for ; Wed, 9 Feb 2000 11:57:38 -0800 (PST) mail_from (hawkes@cthulhu.engr.sgi.com) Received: from engr.sgi.com (babylon.engr.sgi.com [163.154.10.144]) by cthulhu.engr.sgi.com (980427.SGI.8.8.8/970903.SGI.AUTOCF) via ESMTP id MAA35976; Wed, 9 Feb 2000 12:01:46 -0800 (PST) mail_from (hawkes@engr.sgi.com) Message-ID: <38A1C998.CA55553D@engr.sgi.com> Date: Wed, 09 Feb 2000 12:10:00 -0800 From: John Hawkes Organization: Silicon Graphics X-Mailer: Mozilla 4.7C-SGI [en] (X11; I; IRIX64 6.5 IP19) X-Accept-Language: en MIME-Version: 1.0 To: Linus Torvalds CC: linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com Subject: Re: [PATCH] spinlock metering (2.3.42) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing Linus Torvalds wrote: > > On Tue, 8 Feb 2000, John Hawkes wrote: > > > > Please consider this SMP i386 patch against 2.3.42 to add a "spinlock > > metering" enhancement. This patch is also found at: > > http://oss.sgi.com/projects/lockmeter/download/ > > Mind showing us an example of real-world usage and results? I'm loathe to > apply this along with all the other changes I have pending, but I'd like > to know how it looks and what the results are under some real usage, just > as an example..? I take a 4x500MHz Xeon running 2.3.28 and exercise it with what I call a "modified AIM7" workload -- without three disk subtests that otherwise produce 90% idle time because these synchronous subtests saturate my single disk spindle. My system thus becomes compute-bound, roughly 75% user and 25% system. Spinlock metering tells me that 4% of the available CPU cycles (160msec per the total of 4,000msec on the four CPUs, per second) are spent waiting on spinlocks, with the longest wait-time being over 14msec and the longest hold-time being 15msec. This 4% is an improvement over the 8% exhibited by 2.2.13. The kernel_flag is still the biggest culprit in 2.3.28 and it accounts for almost all of the witnessed wait-time cycles (157msec of the 160msec total). A call in ext2_get_block() accounts for 45msec of this 157msec. The largest hold-time occurs in do_close() -- that 15msec mentioned earlier -- with a mean hold-time of 59usec (microsec). Other long holds are done by sys_unlink() (10msec max), sys_open() (5.5msec), and sys_execvs() (2msec). -- John Hawkes hawkes@engr.sgi.com http://oss.sgi.com/projects From owner-lockmeter@oss.sgi.com Wed Feb 9 12:47:08 2000 Received: by oss.sgi.com id ; Wed, 9 Feb 2000 12:46:58 -0800 Received: from zmamail01.zma.compaq.com ([161.114.64.101]:52233 "HELO zmamail01.zma.compaq.com") by oss.sgi.com with SMTP id ; Wed, 9 Feb 2000 12:46:26 -0800 Received: by zmamail01.zma.compaq.com (Postfix, from userid 12345) id CAF2614F; Wed, 9 Feb 2000 15:46:19 -0500 (EST) Received: from flume.zk3.dec.com (broflume.zk3.dec.com [16.141.0.42]) by zmamail01.zma.compaq.com (Postfix) with ESMTP id 8703C363; Wed, 9 Feb 2000 15:46:19 -0500 (EST) Received: from zk3.dec.com by flume.zk3.dec.com (8.8.8/1.1.8.2/16Jan95-0946AM) id PAA0000028803; Wed, 9 Feb 2000 15:46:19 -0500 (EST) Message-ID: <38A1D190.20EB0B61@zk3.dec.com> Date: Wed, 09 Feb 2000 15:44:00 -0500 From: Peter Rival Organization: Tru64 QMG Performance Engineering X-Mailer: Mozilla 4.61 [en] (X11; I; Linux 2.2.12-20smp i586) X-Accept-Language: en MIME-Version: 1.0 To: Linus Torvalds Cc: John Hawkes , linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com, ezolt@perf.zko.dec.com Subject: Re: [PATCH] spinlock metering (2.3.42) References: Content-Type: multipart/mixed; boundary="------------363A90554A9E02044A946C45" Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing This is a multi-part message in MIME format. --------------363A90554A9E02044A946C45 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Linus Torvalds wrote: > On Wed, 9 Feb 2000, Peter Rival wrote: > > > > Not to ask a stupid question, but what type of "real-world usage" would you > > like to see? I did most of the Alpha work (yes John I still owe you an > > update - in your inbox by end of the day) so I have scads of example outputs > > under various loads, but almost all of them are benchmark-driven. If there > > is anything of that ilk that would be remotely interesting to you, let me > > know and I'll send you what I have (or generate a new one). > > Benchmark-driven is fine, if the benchmark is meaningful. Specweb or > similar would be fascinating, as would lmbench or the SAP or Oracle load > benchmarks. > I don't have Specweb specifically, but I know that Phil Ezolt did some work with the Alpha lockmeter patches on 2.2 (Phil?). I'm currently trying to get time on our Specweb testbed as it stands now, so I'll accelerate that request if possible. I'm attaching a copy of the output from a lockstat trace of an AIM VII (don't shoot - it's all I can find right now! ;) shared system benchmark at the (I believe) 2000 user datapoint (4CPU EV67 ES40 w/ the max 2GB of memory). I'll get lmbench numbers if they're more interesting to you; Oracle/SAP are a bit tougher as there's no Oracle/SAP for AlphaLinux. As soon as one is available, I can get results. Personally, I like the output of this one better than the lockinfo in Tru64, but that's just me. :) HTH. - Pete --------------363A90554A9E02044A946C45 Content-Type: text/plain; charset=us-ascii; name="locks.warm" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="locks.warm" ___________________________________________________________________________________________ System: Linux schooner.zk3.dec.com 2.3.34 #2 SMP Sat Jan 11 01:08:08 EST 2048 alpha Command: ./multitask All (4) CPUs Selecting locks: threshhold: >0/sec contention: >0% Start time: Sat Jan 11 13:48:25 2048 End time: Sat Jan 11 14:01:11 2048 Delta Time: 7843.86 sec, slots in use: 398, - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - SPINLOCKS HOLD WAIT TOT/SEC CON MEAN ( MAX ) MEAN ( MAX ) TOTAL NOWAIT SPIN REJECT NAME 28198.67 3% 2.6us(129502us) 59us(129497us) 221186423 216056912 5129511 0 *TOTAL* 65.38 1% 0.4us( 8.7us) 1.1us( 2.9us) 512838 512636 202 0 backlog+0x14 32.69 1% 0.4us( 4.8us) 1.0us( 2.6us) 256419 256293 126 0 net_bh+0xa0 32.69 1% 0.4us( 8.7us) 1.1us( 2.9us) 256419 256343 76 0 netif_rx+0x10c 2342.76 3% 0.8us( 25us) 2.3us( 151us) 18376302 18006816 369486 0 contig_page_data 1171.43 3% 1.0us( 25us) 2.2us( 151us) 9188544 8967242 221302 0 __alloc_pages+0xb4 1171.33 2% 0.6us( 22us) 2.5us( 150us) 9187758 9039574 148184 0 __free_pages_ok+0xe4 32.42 1% 0.5us( 16us) 1.0us( 2.2us) 254326 254202 124 0 device_request_lock 24.30 1% 0.5us( 16us) 1.0us( 1.7us) 190571 190505 66 0 scsi_allocate_device+0x7c 8.13 1% 0.5us( 10us) 1.0us( 2.2us) 63755 63697 58 0 scsi_release_command+0x40 813.58 1% 0.4us( 743us) 1.7us( 65us) 6381626 6377221 4405 0 files_lock 201.94 1% 0.4us( 244us) 1.6us( 65us) 1583970 1582669 1301 0 _fput+0x98 196.95 1% 0.4us( 397us) 2.0us( 17us) 1544861 1544599 262 0 file_move+0x38 205.54 1% 0.5us( 743us) 1.8us( 32us) 1612255 1610226 2029 0 get_empty_filp+0x2c 205.54 1% 0.4us( 349us) 1.8us( 7.7us) 1612255 1611450 805 0 get_empty_filp+0xd0 3.61 1% 0.4us( 31us) 1.4us( 2.8us) 28285 28277 8 0 put_filp+0x48 174.07 1% 0.5us( 342us) 1.6us( 9.7us) 1365355 1365194 161 0 inode_lock 25.17 1% 0.4us( 99us) 1.5us( 2.8us) 197441 197422 19 0 __mark_inode_dirty+0x38 27.42 1% 0.5us( 342us) 1.4us( 9.7us) 215053 214986 67 0 get_empty_inode+0x4c 1.27 1% 0.6us( 19us) 1.3us( 1.7us) 10000 9998 2 0 get_new_inode+0x74 1.27 0% 0.7us( 37us) 0us 10000 10000 0 0 iget4+0x74 18.73 1% 0.5us( 252us) 1.8us( 3.2us) 146945 146935 10 0 insert_inode_hash+0x60 43.77 1% 0.4us( 81us) 1.9us( 3.4us) 343325 343283 42 0 iput+0x68 23.83 1% 0.4us( 88us) 1.7us( 3.1us) 186945 186929 16 0 iput+0x130 28.69 1% 0.5us( 71us) 1.7us( 1.7us) 225053 225052 1 0 iput+0x198 0.02 0% 2.0us( 12us) 0us 154 154 0 0 sync_inodes+0x44 3.88 1% 1.8us( 119us) 1.8us( 2.5us) 30439 30435 4 0 sync_inodes+0x154 177.84 1% 1.6us( 1423us) 5.6us( 683us) 1394913 1387689 7224 0 io_request_lock 3.81 6% 1.3us( 19us) 7.3us( 655us) 29918 28157 1761 0 __get_request_wait+0x268 60.49 1% 0.9us( 733us) 5.9us( 683us) 474446 472148 2298 0 __make_request+0x2d8 11.33 1% 0.6us( 15us) 4.4us( 24us) 88899 88013 886 0 add_request+0x130 7.86 1% 6.3us( 25us) 9.4us( 348us) 61623 61253 370 0 do_isp1020_intr_handler+0x4c 60.49 1% 0.5us( 677us) 5.2us( 13us) 474446 474388 58 0 ll_rw_block+0x1f0 8.13 2% 2.7us( 20us) 3.1us( 340us) 63755 62493 1262 0 scsi_dispatch_cmd+0x1ec 8.13 1% 0.5us( 5.9us) 3.3us( 14us) 63755 63713 42 0 scsi_old_done+0x620 8.13 1% 12us( 1423us) 2.8us( 283us) 63755 63295 460 0 scsi_queue_next_request+0x4c 8.13 1% 1.5us( 15us) 3.3us( 17us) 63755 63741 14 0 scsi_request_fn+0x314 0.15 3% 3.0us( 8.3us) 20us( 306us) 1185 1161 24 0 sym53c8xx_intr+0xe0 1.20 1% 8.9us( 674us) 3.8us( 21us) 9376 9327 49 0 unplug_device+0x3c 2063.67 10% 15us(129502us) 179us(129497us) 16187121 14671047 1516074 0 kernel_flag 11.54 21% 1.1us( 661us) 274us( 6642us) 90534 72350 18184 0 _fput+0x3c 195.27 12% 16us( 2640us) 66us( 10587us) 1531645 1361895 169750 0 do_close+0x13c 15.96 63% 958us( 3325us) 957us(126797us) 125180 47456 77724 0 do_exit+0xf8 15.96 65% 322us(129502us) 1154us( 22483us) 125180 44681 80499 0 do_fork+0x130 0.02 8% 6.7us( 23us) 532us( 5756us) 153 142 11 0 do_select+0x108 0.34 6% 2.4us( 55us) 51us( 2138us) 2635 2492 143 0 ext2_get_block+0x98 447.20 9% 1.9us( 773us) 118us(129497us) 3507747 3193027 314720 0 ext2_get_block+0x38c 0.00 0% 34us( 45us) 0us 3 3 0 0 fsync_dev+0x40 65.89 15% 9.5us( 831us) 95us( 8513us) 516794 439284 77510 0 osf_mmap+0xd0 0.13 20% 15us( 79us) 271us( 2091us) 1000 808 192 0 osf_shmat+0x48 0.71 21% 544us( 1827us) 502us( 7188us) 5551 4409 1142 0 schedule+0x69c 4.73 22% 3.6us( 331us) 317us(124260us) 37107 29158 7949 0 sock_map_fd+0x94 0.38 17% 7.2us( 58us) 456us( 5496us) 3016 2517 499 0 sock_poll+0x94 0.02 33% 1049us( 5074us) 663us( 4224us) 151 102 49 0 sync_old_buffers+0x3c 0.98 11% 13us( 62us) 239us( 4589us) 7680 6850 830 0 sys_chdir+0x3c 0.12 0% 5.1us( 36us) 0us 910 910 0 0 sys_chmod+0x44 0.12 0% 5.2us( 39us) 0us 910 910 0 0 sys_chown+0x4c 3.09 27% 482us( 1734us) 339us( 4524us) 24215 17814 6401 0 sys_execve+0x4c 0.00 0% 1.5us( 10us) 0us 35 35 0 0 sys_fchdir+0xe8 20.26 9% 0.6us( 101us) 59us( 8878us) 158894 144798 14096 0 sys_fcntl+0xd0 39.85 8% 11us( 725us) 73us( 9953us) 312555 288122 24433 0 sys_getdents+0x58 3.71 11% 1.3us( 714us) 84us( 4626us) 29072 25990 3082 0 sys_ioctl+0xd4 10.35 1% 15us( 291us) 63us( 2400us) 81165 80519 646 0 sys_link+0x54 877.50 6% 0.4us( 175us) 36us( 7658us) 6883004 6488031 394973 0 sys_lseek+0xf4 0.11 0% 40us( 92us) 0us 840 840 0 0 sys_mkdir+0x50 43.01 16% 1.6us( 194us) 55us( 7565us) 337380 284984 52396 0 sys_newfstat+0x50 37.11 14% 7.1us( 377us) 43us( 7045us) 291110 251155 39955 0 sys_newlstat+0x44 37.45 17% 10us( 405us) 72us( 11041us) 293732 245291 48441 0 sys_newstat+0x44 194.38 12% 11us( 1566us) 68us( 67063us) 1524716 1355861 168855 0 sys_open+0x80 3.09 15% 0.8us( 48us) 116us( 6871us) 24215 20760 3455 0 sys_personality+0x3c 0.13 19% 25us( 82us) 387us( 3225us) 1001 820 181 0 sys_pipe+0x34 0.19 14% 5.0us( 27us) 460us( 5743us) 1508 1307 201 0 sys_poll+0x60 1.27 8% 139us( 1145us) 28us( 7013us) 10000 9289 711 0 sys_readlink+0x5c 0.11 0% 455us( 2438us) 0us 840 840 0 0 sys_rmdir+0x50 28.97 3% 16us( 1681us) 77us( 6222us) 227270 222429 4841 0 sys_unlink+0x48 0.12 0% 5.3us( 42us) 0us 910 910 0 0 sys_utimes+0x44 0.02 18% 3.9us( 35us) 9.2us( 3913us) 149 123 26 0 tty_read+0xec 0.03 48% 7.8us( 956us) 6.2us( 4742us) 207 109 98 0 tty_write+0x2b0 3.58 15% 9.0us( 120us) 117us( 4606us) 28107 24026 4081 0 unix_find_other+0x5c 27.83 1% 4.9us( 1091us) 11us( 395us) 218289 217653 636 0 loopback_dev+0x118 27.83 1% 4.9us( 1091us) 11us( 395us) 218289 217653 636 0 dev_queue_xmit+0x18c 27.83 1% 0.4us( 316us) 1.6us( 7.7us) 218289 218228 61 0 loopback_dev+0x120 27.83 1% 0.4us( 316us) 1.6us( 7.7us) 218289 218228 61 0 dev_queue_xmit+0x5c 1244.21 1% 1.3us( 5670us) 1.7us( 3479us) 9759376 9729763 29613 0 lru_list_lock 1.92 1% 1.3us( 74us) 1.8us( 2.5us) 15035 15032 3 0 __bforget+0x34 8.92 3% 0.6us( 69us) 3.4us( 3479us) 69979 68324 1655 0 flush_dirty_buffers+0x48 1.92 1% 1.5us( 55us) 1.3us( 11us) 15035 14953 82 0 insert_into_queues+0x9c 1005.75 1% 0.5us( 690us) 1.6us( 67us) 7888931 7861151 27780 0 refile_buffer+0x34 0.00 0% 1377us( 2480us) 0us 19 19 0 0 sync_buffers+0x15c 1.37 1% 2.5us( 5670us) 1.0us( 1.5us) 10757 10756 1 0 sync_buffers+0x1c8 224.33 1% 4.8us( 822us) 1.4us( 15us) 1759620 1759528 92 0 try_to_free_buffers+0x64 4862.69 5% 0.6us( 970us) 1.4us( 820us) 38142240 36446576 1695664 0 pagecache_lock 497.85 8% 0.8us( 639us) 1.4us( 714us) 3905070 3630946 274124 0 __find_get_page+0x44 2258.15 5% 0.5us( 970us) 1.4us( 820us) 17712576 16882375 830201 0 __find_lock_page+0x44 224.37 7% 1.9us( 367us) 1.2us( 78us) 1759908 1651618 108290 0 add_to_page_cache_unique+0x48 1304.38 5% 0.5us( 727us) 1.4us( 316us) 10231376 9793515 437861 0 do_generic_file_read+0x21c 224.33 1% 0.5us( 360us) 1.4us( 72us) 1759620 1756404 3216 0 remove_inode_page+0x38 353.61 2% 0.4us( 405us) 1.5us( 343us) 2773690 2731718 41972 0 truncate_inode_pages+0x54 448.70 1% 0.4us( 358us) 2.0us( 33us) 3519528 3519327 201 0 pagemap_lru_lock 224.37 1% 0.5us( 93us) 1.7us( 33us) 1759908 1759857 51 0 add_to_page_cache_unique+0x128 224.33 1% 0.4us( 358us) 2.1us( 12us) 1759620 1759470 150 0 truncate_inode_pages+0xec 256.13 8% 58us( 586us) 176us( 1369us) 2009033 1864156 144877 0 runqueue_lock 22.53 15% 2.7us( 28us) 286us( 1201us) 176742 150453 26289 0 __wake_up+0x17c 0.35 10% 2.2us( 10us) 181us( 676us) 2726 2464 262 0 process_timeout+0x40 35.61 3% 0.4us( 13us) 283us( 1146us) 279356 271310 8046 0 release+0x44 1.67 27% 2.6us( 11us) 136us( 876us) 13100 9682 3418 0 schedule_tail+0x74 46.07 6% 320us( 586us) 248us( 1369us) 361353 343131 18222 0 schedule+0xfc 6.16 20% 2.4us( 16us) 159us( 1283us) 48327 39111 9216 0 schedule+0x3c8 0.03 4% 33us( 426us) 169us( 651us) 232 225 7 0 schedule+0x744 127.62 1% 0.3us( 4.1us) 232us( 1285us) 1001017 996692 4325 0 send_sig_info+0x2f0 16.09 60% 3.1us( 21us) 110us( 1090us) 126180 51088 75092 0 wake_up_process+0x48 51.25 1% 0.6us( 76us) 11us( 130us) 402000 401917 83 0 sem_ids+0x48 0.13 1% 1.1us( 16us) 1.1us( 1.1us) 1000 999 1 0 semctl_down+0xb4 0.13 1% 2.1us( 22us) 26us( 41us) 1000 999 1 0 semctl_main+0x80 51.00 1% 0.6us( 76us) 11us( 130us) 400000 399919 81 0 sys_semop+0x13c 162.55 1% 0.5us( 12us) 1.5us( 7.4us) 1275026 1274396 630 0 timerlist_lock 19.49 1% 0.5us( 8.8us) 1.7us( 4.7us) 152842 152605 237 0 add_timer+0x44 37.18 1% 0.6us( 12us) 1.5us( 7.4us) 291628 291414 214 0 del_timer+0x40 3.84 1% 0.7us( 3.8us) 1.4us( 3.2us) 30091 30023 68 0 mod_timer+0x44 98.95 1% 0.5us( 11us) 1.3us( 2.7us) 776126 776020 106 0 timer_bh+0x20c 3.10 1% 0.4us( 3.4us) 1.3us( 2.1us) 24339 24334 5 0 timer_bh+0x42c 15.29 1% 0.4us( 4.7us) 1.1us( 2.2us) 119924 119899 25 0 tqueue_lock 0.43 1% 0.3us( 1.2us) 1.1us( 1.3us) 3357 3350 7 0 __get_request_wait+0x1d4 1.20 1% 0.3us( 0.8us) 1.1us( 2.2us) 9376 9365 11 0 __make_request+0x35c 0.36 1% 0.4us( 3.0us) 1.1us( 1.1us) 2833 2831 2 0 __wait_on_buffer+0x1e0 6.63 1% 0.4us( 2.3us) 0.9us( 1.4us) 51993 51989 4 0 batch_entropy_store+0xe4 0.01 0% 0.3us( 1.5us) 0us 110 110 0 0 do_serial_bh+0x40 0.01 0% 0.4us( 2.5us) 0us 110 110 0 0 rs_sched_event+0x7c 0.01 0% 0.4us( 0.5us) 0us 70 70 0 0 sync_old_buffers+0xc0 6.63 1% 0.3us( 4.7us) 1.5us( 1.5us) 52034 52033 1 0 tqueue_bh+0x40 0.01 0% 0.4us( 1.0us) 0us 41 41 0 0 tty_flip_buffer_push+0x80 673.07 1% 1.0us( 429us) 1.0us( 63us) 5279436 5271869 7567 0 unused_list_lock 448.74 1% 0.4us( 409us) 1.0us( 63us) 3519816 3512338 7478 0 get_unused_buffer_head+0x38 224.33 1% 2.2us( 429us) 1.0us( 3.3us) 1759620 1759531 89 0 try_to_free_buffers+0xd0 3.58 1% 0.5us( 2.3us) 0.9us( 5.9us) 28082 28081 1 0 wait_for_request 1.79 0% 0.4us( 1.8us) 0us 14041 14041 0 0 __get_request_wait+0x94 1.79 1% 0.6us( 2.3us) 0.9us( 0.9us) 14041 14040 1 0 __get_request_wait+0x360 3.78 1% 0.4us( 13us) 5.0us( 5.0us) 29638 29637 1 0 __down+0x1f0 3.78 1% 0.7us( 4.9us) 14us( 250us) 29638 29607 31 0 __down+0x88 3518.42 1% 0.7us( 1420us) 1.2us( 38us) 27597965 27487596 110369 0 __wake_up+0x54 34.24 1% 0.4us( 79us) 7.2us( 136us) 268604 268573 31 0 do_munmap+0x258 0.13 1% 0.4us( 19us) 20us( 39us) 1000 999 1 0 inet_wait_for_connect+0x2a0 97.36 1% 0.4us( 302us) 0.9us( 4.9us) 763711 763691 20 0 insert_vm_struct+0x184 0.25 1% 0.7us( 29us) 1.1us( 1.3us) 2000 1999 1 0 ipc_addid+0xc8 119.99 1% 0.8us( 20us) 1.4us( 3.8us) 941178 940751 427 0 kfree+0xa4 119.99 1% 0.8us( 17us) 1.3us( 4.1us) 941178 940658 520 0 kmalloc+0x6c 2576.84 3% 0.6us( 20us) 1.5us( 26us) 20212388 19769826 442562 0 kmem_cache_alloc+0x4c 2575.71 4% 0.5us( 25us) 1.5us( 26us) 20203540 19414071 789469 0 kmem_cache_free+0x50 0.05 1% 0.4us( 2.2us) 0.7us( 2.1us) 373 372 1 0 kmem_cache_grow+0x42c 25.66 1% 0.4us( 5.6us) 0.9us( 1.0us) 201312 201308 4 0 skb_recv_datagram+0xe8 4.73 1% 0.4us( 50us) 38us( 113us) 37107 36822 285 0 sock_fasync+0xb4 0.13 1% 0.4us( 0.9us) 12us( 23us) 1000 996 4 0 tcp_accept+0x6c 0.38 1% 0.4us( 14us) 53us( 124us) 3000 2982 18 0 tcp_close+0x68 0.38 1% 4.8us( 43us) 20us( 20us) 3000 2999 1 0 tcp_close+0x6e8 2.92 5% 4.7us( 680us) 1.2us( 18us) 22913 21960 953 0 tcp_data_wait+0x1e0 2.92 1% 0.4us( 61us) 19us( 397us) 22913 22752 161 0 tcp_data_wait+0x2a8 2.16 1% 34us( 1105us) 7.3us( 275us) 16938 16818 120 0 tcp_delack_timer+0x2c 11.35 1% 1.4us( 715us) 3.3us( 20us) 89000 88893 107 0 tcp_recvmsg+0x4e0 11.35 2% 0.4us( 610us) 10us( 63us) 89000 87510 1490 0 tcp_recvmsg+0x88 15.21 5% 60us( 1365us) 1.4us( 61us) 119289 114403 4886 0 tcp_v4_rcv+0x330 11.35 1% 1.4us( 82us) 1.2us( 1.7us) 89000 88989 11 0 tcp_v4_sendmsg+0x184 11.35 2% 0.4us( 125us) 23us( 86us) 89000 88039 961 0 tcp_v4_sendmsg+0x60 12.62 1% 0.3us( 1.6us) 1.0us( 1.2us) 99000 98999 1 0 udp_queue_rcv_skb+0x94 0.42 1% 0.7us( 2.1us) 1.6us( 2.2us) 3312 3309 3 0 wait_for_packet+0x24c 0.58 1% 0.4us( 17us) 36us( 426us) 4529 4490 39 0 wait_for_tcp_memory+0x2e8 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - RWLOCK SPINLOCK WAIT TOT/SEC CON MEAN ( MAX ) TOTAL NOWAIT SPIN REJECT NAME 7124.46 1% 81us(125881us) 55883246 55847811 35435 0 *TOTAL* 27.83 1% 0.5us( 1.2us) 218289 218255 34 0 [0xfffffc005fc2f480] 27.83 1% 0.5us( 1.2us) 218289 218255 34 0 ip_output+0x78 0.81 1% 0.1us( 0.2us) 6368 6367 1 0 arp_tbl+0x134 0.81 1% 0.1us( 0.2us) 6342 6341 1 0 neigh_lookup+0x80 0.00 0% 0us 26 26 0 0 neigh_periodic_timer+0x3c 5.42 1% 0.3us( 0.7us) 42537 42534 3 0 fib_hash_lock 5.42 1% 0.3us( 0.7us) 42537 42534 3 0 fn_hash_lookup+0x44 796.44 1% 0.8us( 19us) 6247150 6232023 15127 0 hash_table_lock 1.92 0% 0us 15035 15035 0 0 __bforget+0x48 568.27 1% 0.8us( 19us) 4457460 4442363 15097 0 get_hash_table+0x98 1.92 1% 0.0us( 0.4us) 15035 15034 1 0 insert_into_queues+0xb0 224.33 1% 0.2us( 1.2us) 1759620 1759591 29 0 try_to_free_buffers+0x78 28.04 1% 0.2us( 0.3us) 219968 219948 20 0 inet_protocol_lock 28.04 1% 0.2us( 0.3us) 219968 219948 20 0 ip_local_deliver+0xec 3.33 1% 0.2us( 0.7us) 26147 26140 7 0 inetdev_lock 0.81 1% 0.3us( 0.7us) 6342 6337 5 0 arp_rcv+0x48 0.00 0% 0us 28 28 0 0 fib_validate_source+0x60 0.00 0% 0us 15 15 0 0 inet_select_addr+0x50 2.52 1% 0.2us( 0.6us) 19762 19760 2 0 ip_route_input_slow+0x4c 32.69 1% 0.6us( 1.8us) 256419 256227 192 0 ptype_lock 32.69 1% 0.6us( 1.8us) 256419 256227 192 0 net_bh+0x1a8 3.46 1% 0.2us( 0.3us) 27104 27091 13 0 rt_hash_lock 2.73 1% 0.2us( 0.3us) 21413 21401 12 0 ip_route_input+0xd4 0.64 1% 0.2us( 0.2us) 5000 4999 1 0 ip_route_output+0xdc 0.08 0% 0us 663 663 0 0 rt_check_expire+0x7c 0.00 0% 0us 28 28 0 0 rt_intern_hash+0xa0 239.74 2% 151us(125881us) 1880507 1861435 19072 0 tasklist_lock 0.02 0% 0us 154 154 0 0 count_active_tasks+0x2c 15.96 1% 82us( 755us) 125180 124604 576 0 do_fork+0x8ec 15.96 1% 14us( 644us) 125180 124536 644 0 exit_notify+0x40 15.96 1% 11us( 1187us) 125180 124266 914 0 exit_notify+0x244 0.00 6% 1.0us( 1.0us) 18 17 1 0 get_pid+0xb0 0.00 0% 0us 1 1 0 0 kill_pg_info+0x60 127.49 1% 0.8us( 30us) 1000000 996102 3898 0 kill_something_info+0x120 1.27 1% 1.0us( 3.9us) 10000 9976 24 0 proc_pid_lookup+0x90 0.17 1% 1.2us( 960us) 1358 1352 6 0 proc_root_lookup+0x6c 15.96 6% 404us(125881us) 125180 118256 6924 0 release+0xa4 0.03 0% 0us 232 232 0 0 schedule+0x6d4 0.00 0% 0us 1 1 0 0 sys_setsid+0x2c 46.92 2% 1.1us( 987us) 368023 361938 6085 0 sys_wait4+0x1f8 13.60 1% 0.1us( 0.2us) 106679 106676 3 0 udp_hash_lock 12.62 1% 0.2us( 0.2us) 99000 98998 2 0 udp_rcv+0x178 0.25 1% 0.0us( 0.0us) 2000 1999 1 0 udp_v4_get_port+0x58 0.25 0% 0us 2000 2000 0 0 udp_v4_hash+0x64 0.21 0% 0us 1679 1679 0 0 udp_v4_mcast_deliver+0x48 0.25 0% 0us 2000 2000 0 0 udp_v4_unhash+0x50 15.35 1% 1.4us( 24us) 120428 120353 75 0 unix_table_lock 7.68 1% 0.8us( 1.7us) 60214 60191 23 0 unix_create1+0x17c 7.68 1% 1.6us( 24us) 60214 60162 52 0 unix_release_sock+0x40 202.83 1% 7.9us( 730us) 1590980 1590646 334 0 xtime_lock 3.86 1% 30us( 730us) 30299 30235 64 0 do_gettimeofday+0x3c 98.95 1% 2.4us( 65us) 776126 775943 183 0 timer_bh+0x40 100.02 1% 3.5us( 272us) 784555 784468 87 0 timer_interrupt+0x60 15.96 1% 0.1us( 0.1us) 125180 125178 2 0 copy_files+0x118 201.44 1% 0.0us( 0.1us) 1580075 1580057 18 0 do_close+0x4c 200.67 1% 0.0us( 0.0us) 1574005 1573996 9 0 do_close+0xcc 205.54 1% 0.0us( 0.1us) 1612255 1612214 41 0 get_unused_fd+0x30 6.17 1% 0.0us( 0.0us) 48430 48429 1 0 open_dentry+0x108 45.39 1% 0.1us( 0.2us) 356032 356025 7 0 osf_mmap+0xf4 38.25 1% 0.1us( 0.2us) 300000 299996 4 0 sock_def_readable+0x24 22.95 1% 0.1us( 0.4us) 180000 179997 3 0 sys_fcntl+0x54 39.85 1% 0.1us( 0.2us) 312555 312548 7 0 sys_getdents+0x7c 877.50 1% 0.1us( 0.2us) 6883004 6882931 73 0 sys_lseek+0x4c 43.01 1% 0.1us( 0.1us) 337380 337371 9 0 sys_newfstat+0x70 190.78 1% 0.0us( 0.0us) 1496431 1496416 15 0 sys_open+0xe4 1354.29 1% 0.1us( 0.3us) 10622897 10622797 100 0 sys_read+0x4c 2320.47 1% 0.1us( 0.3us) 18201471 18201261 210 0 sys_write+0x4c 15.21 1% 0.2us( 0.3us) 119289 119263 26 0 tcp_v4_rcv+0x1b8 26.81 1% 0.3us( 0.7us) 210289 210277 12 0 tcp_write_space+0x24 12.62 1% 0.1us( 0.1us) 99000 98996 4 0 udp_sendmsg+0x24c 12.62 1% 0.1us( 0.2us) 99000 98999 1 0 unix_dgram_sendmsg+0xb8 7.68 1% 0.0us( 0.0us) 60214 60213 1 0 unix_release_sock+0x68 7.68 1% 0.1us( 0.1us) 60214 60213 1 0 unix_release_sock+0x7c 12.62 1% 0.1us( 0.2us) 99000 98998 2 0 unix_stream_sendmsg+0x264 12.62 1% 0.1us( 0.2us) 99000 98996 4 0 unix_stream_sendmsg+0xa8 28.83 1% 0.1us( 0.2us) 226107 226103 4 0 unix_write_space+0x24 ___________________________________________________________________________________________ --------------363A90554A9E02044A946C45-- From owner-lockmeter@oss.sgi.com Wed Feb 9 14:25:21 2000 Received: by oss.sgi.com id ; Wed, 9 Feb 2000 14:25:11 -0800 Received: from leibniz.math.psu.edu ([146.186.130.2]:63206 "EHLO math.psu.edu") by oss.sgi.com with ESMTP id ; Wed, 9 Feb 2000 14:24:51 -0800 Received: from weyl.math.psu.edu (weyl.math.psu.edu [146.186.130.226]) by math.psu.edu (8.9.3/8.9.3) with ESMTP id RAA24053; Wed, 9 Feb 2000 17:24:44 -0500 (EST) Received: from localhost (viro@localhost) by weyl.math.psu.edu (8.9.3/8.9.3) with ESMTP id RAA28601; Wed, 9 Feb 2000 17:24:43 -0500 (EST) X-Authentication-Warning: weyl.math.psu.edu: viro owned process doing -bs Date: Wed, 9 Feb 2000 17:24:42 -0500 (EST) From: Alexander Viro To: John Hawkes cc: Linus Torvalds , linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, lockmeter@oss.sgi.com Subject: Re: [PATCH] spinlock metering (2.3.42) In-Reply-To: <38A1C998.CA55553D@engr.sgi.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing On Wed, 9 Feb 2000, John Hawkes wrote: > The kernel_flag is still the biggest culprit in 2.3.28 and it accounts > for > almost all of the witnessed wait-time cycles (157msec of the 160msec > total). A call in ext2_get_block() accounts for 45msec of this 157msec. > The largest hold-time occurs in do_close() -- that 15msec mentioned Almost definitely dput() work. > earlier -- with a mean hold-time of 59usec (microsec). Other long holds > are done by sys_unlink() (10msec max), sys_open() (5.5msec), and Ditto. Interesting... > sys_execvs() (2msec). Most likely we are dealing with dcache stuff. Sorry, it will have to wait. Sigh... From owner-lockmeter@oss.sgi.com Tue Feb 15 11:53:54 2000 Received: by oss.sgi.com id ; Tue, 15 Feb 2000 11:53:34 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:54654 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Tue, 15 Feb 2000 11:53:30 -0800 Received: from cthulhu.engr.sgi.com (cthulhu.engr.sgi.com [192.26.80.2]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id LAA07987 for ; Tue, 15 Feb 2000 11:56:22 -0800 (PST) mail_from (hawkes@cthulhu.engr.sgi.com) Received: from engr.sgi.com (babylon.engr.sgi.com [163.154.10.144]) by cthulhu.engr.sgi.com (980427.SGI.8.8.8/970903.SGI.AUTOCF) via ESMTP id LAA89555; Tue, 15 Feb 2000 11:53:06 -0800 (PST) mail_from (hawkes@engr.sgi.com) Message-ID: <38A9B109.3B4308C9@engr.sgi.com> Date: Tue, 15 Feb 2000 12:03:21 -0800 From: John Hawkes Organization: Silicon Graphics X-Mailer: Mozilla 4.7C-SGI [en] (X11; I; IRIX64 6.5 IP19) X-Accept-Language: en MIME-Version: 1.0 To: linux-kernel@vger.rutgers.edu, linux-smp@vger.rutgers.edu, torvalds@transmeta.com, lockmeter@oss.sgi.com Subject: [PATCH] (2.3.42) revised spinlock metering for ia32 & alpha Content-Type: multipart/mixed; boundary="------------260F86CEEBFC71FB4A0D3738" Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing This is a multi-part message in MIME format. --------------260F86CEEBFC71FB4A0D3738 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I attach a revised version of the spinlock metering code that I contributed last week. This revision fixes some shortcomings that others pointed out, i.e., reducing ia32 bus locking during spin_lock(), and returning proper error codes and length when doing I/O to /proc/lockmeter. This version also supports lockmetering of Alpha SMP kernels, as well as IA32, thanks to the contributions of Peter Rival. An updated "lockstat" command is also available from the website: http://oss.sgi.com/projects/lockmeter/download that is compatible with the corrected /proc/lockmeter return values. Followon work is currently in-progress to implement various enhancements that others have suggested. -- John Hawkes (hawkes@engr.sgi.com) http://oss.sgi.com --------------260F86CEEBFC71FB4A0D3738 Content-Type: application/x-gzip; name="lockmeter1.1.4-2.3.42.diff.gz" Content-Transfer-Encoding: base64 Content-Disposition: inline; filename="lockmeter1.1.4-2.3.42.diff.gz" H4sICKuZqTgAA2xvY2ttZXRlcjEuMS40LTIuMy40Mi5kaWZmANQ8aXPbRrKfoV/RoWMbFA/x kGRLWjtmJNpmosMh5TjJJoUCgSGJFQjQOHQk8f72190zuEiQprP7Uu+5XCQ0M93T3dPT1wxo O5MJNC7NOADX8eL7RqfZbe539s58K54LLzIjx/f2Tn1v4kzjQDRnwl0URjbmIhLBxvE7jUZj e+zadSzgtRhD+wDareN297h1BJ1Wq7VTq9W+dGrtdeAwMtiH1tHxfvd4/1Aie/UKGu32/kGn fgg1+dDuwqtXOwBwIx5CMAMBtkItbHA8KE4UPoTBx2Z0HzWxw3saQWg+wM8Qe64IQ8Ly4McQ CNN1H+DG8+/gbmZGEM2cEGamdYO4RdjcgZ3a9yLwhAuuj43MkeNNd2qnV5evB2+M86vT7y/6 1/3hTg1gxDNEPgjPHLsC6VyFrOM8jjUD07ZD8G9FMBOmTTCji3c8MqwTqnEcAZLm34VMJ/aH QsCtGTh+HEJIPIaRY4UQh4gUyRYSGDvA8udz07ObOzXCNJgwBptlkGPUxyaYiUDUWTKXNBwG o7NLCOMxCi8S8x1QXFIziYy79coAJT4NTBL7SAS3jiVCOHOmTmS6cCmiOz+4CSt1sJABHDK8 HIxoeV4HpmeJ6o69RqUvzBsxcVBsJTqU9K2qatKzqpaoLpvUMgVcUsHnx51OpoJHnfpzqOEn oiXlO3337vV5780Iai+gcWYYuGqGsYMLjlyhsjROZe/xC/haT8ZWofEBRYGfYRQ4VtRYBH7k Rw8LFFvjqgONiT93osYkMOeisfAdlG6AClaOaboZ2Q70yuBIFZyJLSawqrfvR33j9bB30X93 PYQX0N6pSW4QwCOIfH9KFrFfSnYiCXiEqingaWPi+Q1Fquk6Jqnr0zrrt+/h1nMmrLyoswtc jACVxoMIlwacCKWdTva1HqK9cGk4snWKgijBC40RNHzYs8XtnheTkO6t3F8vs8fOyyftE5rY A2HNfKiUoaucwMRZr69mYM32THcxM/csNmpNVPISRSsbt6rHZaNKdHp/o06XIvmAe/A7lGrn ENr7x+3Ocfcg0+/O0WF9H2r09YwVfOLQ4o1934WnF+bUsWD0EA4/ks19mijPRe/N4NQY/Twa /kBaAv+EyteqC/dDBZWo8lCB36SA2ZoxOmVIz8kcXihz+LREIYkEVCM05fFm4Tvd54dbyL44 bI3oi4NKJH/4eckv4SgaloPWcaebE/yBFPxBInjcMVJMZ2IcT+FmTubf2ruZBEKkUjrrf/ue 5H+O4vrfXaWC0/rPV0m6Qn42bsKHedi0NouxBGDTypUML1nDZ1uuYRm2pdXsHLef5yKV/Xa9 i3EKfrU7vJyPUkPY/+nd1fAaF+Li26tzfSoi486amV71hIzyI1wYNLKOJ2w9W5wqPHmy3JyK Ho15EaURLhzPoAUzCGlZJ0Y9m7qj4GFdP0ZI9lrcd4ETic/05nEroaxTl0m4h97M4g9j7oRW uZKUDFtVjZJBywqxf9xqb1SIMhx5a9ola9o5ytTg4KhN4Sp97bdYDbRARHHgoX+LvehkBz5R RPlonSsW9zitB2Ho/I6CA9IVkh1TYzjexNdRcwLYxZCNR9Rxm04mOHKXpLsEvYg/Ay01cG8X 9/4uXFPgO4k9i4JnMC2M6UIMKQpmAAhJMOf4ugkEtrdT41jUSmdN9YWBdHSpsRUBx3W79Ilh IRMxjicY6GqaBvxPQbOYMq4WCz9Edf8DByo5loiEMCWADIBsfcpx9gHVkGNkjAx4OVN4JDYU UagCkJgCmLCcq0zRS9gCxZfvhZHiDpbY25q/kkXL+Es5S6jLqDD8hQh4YUJgdc2Q5HpeAE11 +f78vA7Zv71dcDG7uGHGteICEhNLzGPTpxNJBU/veL69af6VfknFk7VkqgmksWDylncKU0r5 yhaCwI/ldiIANooBpBgUbJ0392GbN/fh/kEdwzDe3CNj8HrYfwN/Aj4N34+GdWjXoUX/sRe/ JZM3lh+IFTmgMTjZYAuKrDEe2wkMzG+DB/ln4Ps5ZVFixUmP6lBJmyu0gkt0vrmSTx+KFNcy itevnlwbZWow5i/ZkQYpu37rO3bO5sOW/CiRqzVCgp4hN6qxUl+V+RpeeMGetSnEetbuYg7H 66XJecQUk2fcxU/SaRO+8yQg8SlErrlJ3CF5ertGLYYrvCpueGr1J3rshc7U43pExAg+bVji 7chJhUvS1EraE4rWrUR+FT6tdb6OZ7mxLfbMcN6QKUQ2wazMO24EWHXIG4ezWz0TFnTb0D48 brXwP7SPDo/WuebN2C7QhbGj30dff9wtxmqtegsjtfohO+jESbCXiDAdHD/Ad/7Mg7fm3Q36 P33G36/CqdPEHLUqB39rhkgvTnLjklqjfAiMqkajSGDQFgB6CH4oAErgCx+XwEF4hHnHO3fo 3Jou6JOAvl/9ftNt2sJKofZUnMjJuNE7f/e2lymR8Rb7ZKRY2keQUlTwD5TV3uwuWIybs5cZ 1Om798bpz6fnlOT3f3jfvzz9WeNRjZfWg4WGExOQj0VEvB57t+gtKRBgZGiozweX738yfuwP R4OrS+P06qwP/4Dv+8PL/nnSqnfq3Xqrmk2Oi2a6hhN8NELzVuj3VfgVddzgvwzTsw3LdbC1 DAB9d4RmNYNRDcbENaehBGKl19CwnxPFoCiGTrPTvE8EqxCP3g0uSXIqbRpcDq7RI8gxytoZ uKeQdWEYSAYGFJ7vBcKKEScSy/5Cxbp6cMchL3p4+lYenmEiMV/UkdLYjdiHIt24LIjRMG59 Fy0kCtzQsb2iubZruPC4XX/c+tWrcNPcv4Wvu9jSUS3KdY3dcUgj25NiexiPGUMng9DCyDUs eNzJ0LaPtflYPh9D5cW8Arph2PF8/sDmRGceqmiGXzwJsI9YyP6SvFQl7BpQ7GTbpWKcBCaJ ZrYRbyGX+Pvkq41RNfKC1fyAR7T/T0h0vUxV8ISqv2oVpOZv7wcW04jK2dt6gXT4Nj4gHVyS nG2u327CVSx5tdrH+51iyetZvuSVmADK+g3TtgMDzbBj6/RY1fR2lYszapDj41afmwtjYU5x 15veVKCtN4OIapohgcj0qk4xTUS2qaGtA7h1gsiIfGMxewhxlR2fB+oKDylIDhOv9V/CgwHR EqocPwvcVv3h8GqoCyIWMMDA0OVGrzwOjx/bxzA2bRoDj1vtQ/e+iUpdx730eoD+wqAntPr8 RHhQbIhlfcV0dcmo+kAava1uZeO3Ua5s9H9Bu3LINqrXAR0YHKiIc9lnysISuUwo+mW0cwGG wtyz00gWR5b7EtcE7eWe4QfVXlsD0VruURAtUgEMfQADkpEzX2DATfzJdD8L9psA13RExIdt 0Z3PB1CmF4V1DH4wZ3aFGYQwGP7wNJS8P+fyx3O1sdLsV6dJq42XjP4bqFCB36sA2jeqbtqo UskI3zOsRVw9IVX8REdlSJmOEQM8EpicZXHQSjCd5BfShwBlH5ArfenJ8pHTAOk1OMOYjyk+ JuKfH5FZOGrV99tJQVaJLl8jkwYY9K8iDDc4RAkx7B47kd6qK+NcXYLkaf07j3msA372rjEY QtMikyRkr5HZ66UVVFEKcb82E1YOlHnOFQMZu3SRxiLi9OERLBOm+MmDKZ9aItCCS85QZPNU KQdNEmrpkOeLk6wJOXiPjGBc0giFJbPxyKeqBxdhTCuKTao9+wswZRPKJQPHf/5EntH6438J lfI/RRwoAlAYIfTBlGhFMMG2AvyYTiFnaAeF7ViykMUyTPrXxgvo02XEAGgL64/baYRAsZds 7OSjBFCN+fhCRgrLCMRHWEGQhA+VZsJVk3js1H+tmPe/VlRnh2lKoMtpGqeNiqZ2ihqFcEvn yUmcwrGICkXWBx4clZT1LQd0OX1U9ecyjcwrpNqreZVMIL9EKUu2/Ipyyn2f01XrRtmnF9CS jKzGpVTvKFbMFUcrnBStxRKM4iXb81+VbmvZX2oS2Hr/9X+EgI8tWh2yee3us9RfTQqOYgfy a5kVAnMRuDSmJ8WRaRGxZGAjMeWJTSt4Jmb8S4z8BqKkhWczFIjp/Yk6qtnnKyXPW8mZ/qY4 GyjOBhlSw0YjvKT22blJjqzgLq8oeSFVlyCqS+hyJy1r8RUkUV2GqRbTglKVg2U3lC0KbBC+ 2m8r4t8mDuTzti+oNC2P3xwHLo/+j+pMK8jyZSaMA5+l50m5MhPGiPky06m/eAic6SwC/bSK Mx8d1QkCgzDXsRDdm8BczBwLA6yBZzWzotHfUZ4qFpoGdPhZXmda7Uq1Na1/8q7DUM6Y/X6y udgkB/2tFaa1jr6yiMPZxIUTWPgLF/0nPlmuUzmuvJiihbivHtPe8HzcBBSz7O7BcWUu5n7w UKnmq0nrylSbJ1YT4tQTt5LM5MdROpWiITfll5S4CkU1rpJVNW6TFTPZwvzqS8ZiZcW3LyHw pgkxw9gqyyuO3mJvp2NLMrzNp/0bUC3t68PjTq58vN8il8mffBuRbONqpMDWkc2hTPlkpE8O qUGa3JYK3YY//9xwzg8ajtG/qtVUdigDFMMWi2hGvVoWgj+R0/BSkseSPlblZM9yOdkfCmsJ StxMLUaLll9nR/2FlDYam7DmQ7NycrfXp63LBkvDt9GoTUWDL1WpDNdGnTrkvBk/O51CRSof AdPlNDSpVKLh0z00FOMocOFrDPxblW2uFSznh5SYrwbka4P3DYMpLl4fGGcWaE3+KINkNI+l ppHL+xnVqRwwE1mbplQ3z1rIMz4/7/IKfGbmzG6WB4qQSm2pVFGetfCepT2YHlFwbMaK8+x5 /QhqR/vJ3eh079LBY3IHacMB5Eraf6LickRQngnnrY6SCLVtEEjh9BHWc78ua5P3cPliVa3d xohK3axh25jlbXnLtTXnhexya94L+vBl3Dc21ZT+gA1VpRP4VN3A0Hq82rrsU2Zhf7XGlQs9 ynOJtMA4xMSmwZkCJIuL8S3fbqf9NI/dyKEKJCVAdAOHs7TuIel0rX24X+90M0+Ls1FVKMED 9BIBwnCB0vFg5QpfSJXwGQZgzZEkqratnczSMa5gLidd67KzsrGJNpbqflmqTNnaml0v9T64 a7yc873Pr16ATNLkweWX7oJi0lnYAePYcW0jT14dKkYOYmLivrAr22zw0iSduOS1PmjLtX5+ VO8eZpv7v8bkUiJcwmWBPmYzB7OOz0Ien+bAVU1D2wGJ3dATZ03BAcb3FSgzFk8IUNmxfP14 Kb+ulqM2bRvjgArKiPk2vh30RsboeggVCg22me/zm0KtJnn6Lzz3NiN/7ljUyJfY6OpM1lSl gSfZKFtYenrXTQaq9FfjpbqfBy8pmqxpyblnOweLEs7BqgGqlpduwS0YpVimeHf1MyWXjP1q Mu0qgmouS14nzM+cciutXUXyWTiptRT7b7Macvup5jAeswsgX6DnNSy5lPj5Oo/M5rcs8qwM Xh+zrwz9y+WdVUzb1Hbah0f/T4s7spJSXt0p6UtZ/NF0Y6Rs4gfshismn01UYGHSW0C0Pxdm SPRFPjB5RrywzUgws9ovv/wCDXrt7A53jonKZwLqbCOM+bYx1xfjkN9J++abb5L7xQld56Pr 3rXRO702Lq+MD73Btaa1ynrJX2qa1i7tO++/QzitU9Y57H/XP8XeblnnRe8n48fe+fv+SNP2 cwIZLYTlmC7cFgWDoQ10ACOmkA6sTA+GvUwyDLginRVWhz1j+GEdN9g56l/0JC8pNad0BZnO Z+ndK5upYRdB1wDkfW6MkZg+nt120PREfvDAwBFVtWzHMjlOK7DlT9I3DBNczaX10SRdF+/P rwfSRvTOzob90Yjv2eZovDDvnXk85/sIecREUQh8gwOR4wZBzxIKedjHZPCE9AIYSdXzIxTv FITnx9NZE/RLH0diNEn3FWbmYiG8MDeWQbPxGHxGM34I/bmaHu4c15V4Bcx9z6GynV2+NKQN /DC4POv/hGvARQFiURI5It7QaMLcDG9SpZiZ4UwySEbb57Z0CconetsbvTWue9+e943R4Jc+ 6l7r6FCjmHseh0xpZ3f3cjPgRW/0vabppQgbbXZLyeJo/BaARBSS2srbuYnSqFXKEQ0DutPu Rabj8c15LfeGAJhjn95ilWozFhzrF+Sq9TJUaip+JzDBmCmdY2O3MyHlJMwNFrPG/ickWxiH 5lQU1TSgCz2LiA9t7vgSA+UHdCbcsBYxw+PeM9WF5JguOPgwdzxnTotnmdaM30QVgKnGFP8n C0SvWpINVReZKcKhEFfb5dddAyMwT3iBcMcjPekbFRG9dItplKSOb9pLsKScIqHyW0yOikNM XyLNQ69v8K7FTSruTzR5lC4trePdZGmU1Gaeb2beCpkvoXXO65/cUIT/07ItkDeyjSj3cgiu E2Nv2M4c9xWyg/OaQWA+SDuDJKglA7UkmeVRL5I0lXoJBYdq1s3Qme4x92sNOiEAL56P0QnD 90IsYBL4c+QhQOJJg7KFCVGnojuBfhNhwgT+bEmhWFqop0qBRKZUCURPXQMYKOMXyhekbxCS VpA4I1hcQ9/i+iZ7X9OT8KBYBLWhudyXuHZmmClQbqCwnzLdd8JslUz53PFsCcl4lGYXwJWo WZ4blDNGW3O4b0SaFc+NO9OJjMhBBZHqhm0xJRO3dCdjTkTSiETz8pAzH9OkjZA0IgfZ7SDk 3LxfmRPbeBZ+Jwc1oARkeTICobY1ICyJf5Y5699OMg3njSaFprQ7kVZJvybb6Gwqafpnme3/ Lf8OFe5HeqfoI7pMesFe6nQDOCFIX9nHdUM7IY2C7+35k0kdyEH48m2kUjdw9fr1arBzdbka Flxdcmg0Ku15e3V+NlqNfbCnd36eOIgEA/y5BFgoqcoudeZGk5WqnrRcmpKlOpqSK5qcUynF ZkvMi1qEcUJD/paCzQaP1V4eMbBABe0cmXHhrmR41AmERoXBpVPac8nmhPcyttHVn5AQ5Iaz caWraPxKSmZg061LE9MgoEErTosxkV4ahEwRaFBDslckyay6uHtDYWG0i7mH4OQk02aCpjv3 aJ4VJP0VktMiIN9TTBZOWTVNicj4F+ZhaOGYdPWMtkx4eZW7Q/uixpfiSqjP40rol14DxZEg z9yYpu2qIx2u2PA1WoZNfBlC5QeUgmKamgEqu5uDBAmp828NsMGtIESlKnGtblh0ydlf7GER rfqhgsRcx7Rlx/Fkgl/ktKQysaOR5jWHe8VFartZU+rBt54AYfmFyoIbpuGGMiHKSqUlC8NQ Z93G6nsdudT15epGHfZ0r6rRwYW8crhb1UHXCytflaUwtMqGJNFQK6e3qvAE/t2tokXwqlAw A2eDIQWUemDSZUZ9GSW1IwSClwal1ZqMPtM3UFXoJgsTMh+gMI7iFkSZ/dKJjLQaMsJKnK4I 6mmkz3cJM8XLwsgoFO4EdNyDfmDLRfIcS2DMGS6okkWn7WmsLvMNigpEWG1imi5/piJGs1Mh b8T7uZL/MRZ1Z5H3uZpX1u7qjIsyAkzW+WcnNvBBRo3k10B9mfpUGWgmAgoEtok6JdBzDLc4 5MIQln4rIIlsMTYKozq6y4WYxPTjNhTimkgCTsJo1G7GnplwbfWbM+pXaRBvKKiITqxI0iU3 9HMxpF5zWnaaDcn9n/aOtbmN2/j59CtgTUclLVKWnabjmJIzTqJk3CZyajutMx6P5kieTI4p HUMexTKO/3uxDwCLxx2PiqdJH/qQ0CSweC32vQveHyfwr4FIGnkw2ihaxNNLaIaSOap2sj8s RstI/5zrjSrGzshPx9ajjYNB+JhRWagWKPWi/CRUStu5s8xvDOLonYYzMGfUpToka1pJPlti yKhG/iEfz2jCoRzTagOAC/2NCrWGSe50IJUjKDN4T1MDFB6xypBWOimNGuJSYdHAAIaFFtKn 5aJe0Xzy5d9+ePr87OLs/OXzp2cv1CcPjEg0K67fVhOzYgoKUft42vts8anTVkK9Q9yWQEGh lpoqCf3mHuHyelJq6fen1RRk4mkVSGX8ywUxRhLmbEd9kLoDMCMLQJJB05cMmYPalZAhXCHr dEzc9CZGbum3DzST/3xdu91SgnRwL0tPQVLfke6j2UI5E7hh1CA01Fl1zZcHlywFWgoBaqi5 20JF0GM2SPkkPKBMNGC3NS8z42nhOumwjdKmeY3GTLg0hidFs2KhTIBzLK8JoMYngHddrFFG Qr0rsHi0EcCczFWjOX10caiJ1evFAqxwdCOy3gWmofVr2hOhbQMaaMyBomKa5y9H+fWyQWbB T6/Pn19otVajH8B6gjqzvpssXJjodqT3xCgFPB9Ns7vy3w6uqoNr/xAk/em2djwG10c6QYMr 7/IahDM3REGpLa0lGmLp2DHaI9FIur6WJhcOAgYcf/Xq1cX5s5c/nr3UX80X+durHM0o+jQ1 Y8krjVtDYDTB0MpHU017NYac3n/w0HMFWQ+2E61swoWNfYvs4dCkncsDk7Uvl20cHrbpNneH bRiHJ5ks51bODgcnCk369IGswfeQKvA97N2nwAtTZqSp+kN5NV3EtQMG7TrP56MLuEep/i2K p9yuKoXnpufABexSVZujEUUf1B06xxo0FbALmsSHHDSID9dUpas53LC/X6fm04f6XEUKJFay 0P/lI3128eyLv2CdtdV0fP/PR6WraHc4vSx+Up0/xEF/vY2W3l1Pp4GUtoLcngp6//W7Z19h RzHku6ty7EbcssdulGRpoLhV7U6LNju7DRMg5CV6+OjBZ4/+dBz5Cz/97HfgLxzt5C/0A7Kx zCCHY/u/FIvFdZn8heqYJX9aaoI/Tv9yNU9+X01AS0tPwY8WD+DZQMzHoeZMpS4TP6zIsGsU 6gBincL95MWLs+cvO2CrEVHow5+LRQmhUz3QUbpZpoV1zcXwm2P6jnjlSX/3PzgrYNDyT96T +j/q+fiWY9psrID3Zt6/oZKLtWL21PvjVKmFD1RRiEk41lC6KCDhbSCGYUGLoF9pQucKyHC0 aU/JqFN70Yx9t3y3muMX+J9vpjcFaM/Pn/RQyauKQLYkkdO5dNmwfy+c0N6hHKATeXZ6sYuI Q2Os6BhoKmnxc04Z/2SUPPV3/Yhlz9fGJOOGAg0GpPZ51AdEWqjvRFm20OQ1An9zZHtD6JUD haoGBofwJE7VMX/JgS+d5OnYnRCwMEgHyuNkdkVyBqG/S7f7QMu/9KdqYMNMwdUMpfNwQnWN zOeuopkn2xEsNyqvDpvZPMcAwfxFk0OeSuBqqVorRBrXQB8io4rzFsU+4DR6BXsao5nSi65F tW2IFeBiT03svgfqFeaJECoahyUaVFDVWU/BsQ5l4DimabbRn6FE84KMMFrmt+XkRloEIy1j XKr3qfwUGB4pVU3k+EGE0ELoh5BZSiHvHHO1jIlBtfiaDLZeLerc9joRLKjjuP1etb8EfAHc 9YMR/GkkdGl1kgxcMPc5ueoEmMNDvLPRquPVicU1dxEXzhqkGjsEW9LmsLLGFhpCQGEQGWuT BragnEFYvxoRY7OHjmmawncfLrG5+BQkpFoxlh6bVCgwy2czdpnQQGvyA9s2MJf5SXzmvrew LjCHYhmI5mBxz/fMdh0y+xtKfj2N2lZoIDSlTYUdE7zAQwHmvylmA13AdnGqtGQJldxAsis1 Wo8xCzzr3PUnwUYY3eVN12I3eplpUW/wKvD8Q7eHuwYH6pMu3VL4OzjgZSRXfBD6W2UH3so7 p1G8WtftjqbGX2th5bpc98g0rhF56Yp7sKEfIpwWLiKfokWQCAMM7zSjy+VZkmBzaBOgI1ya mZ9r6SO9YqyHxrhrHtrpobyh+49rzLlIK0l6Pak3kuOmQMt77Db541ItyvLKGBDRRo6iXa73 QzeQntys3UyYGgaNPauy176e2O0AIjaqc+E/8hQTNu86J2ncD0k3bPSH5BmbzNb0MUuxaStt w+XchsD1hMNB4Xb8N5G87HdG79p188KMwMDTMYFEXfICGbLZApgfP6T31ZwxkcXdIZwqM4dd iLeWsVoS7f9T4/9wavy/RI8lIZ5eTyvy3VGquiI66nI9oUEo/ErXYldCbSpDRYDpS9AHL+Yj deoCVY4FmUwlIFo47qpJViLIIQPvxakDtCNUz8UAsWwEi+rx+SUOE9raK26kfYxrGOXwxA6f p4gBwLAFLK+mydN0Zn1fgMQOAAcmXCvQ8+1FyCsoDonP9RjA0+pIqe84UNn1h4s0XFHFdgi3 zZwPDz4wMmzf0Ey2DFsIDEzsPd6C5gMA258sjC53V/W9vTdoi2PVVl4K0AraaSSHALq2SNah 9pAJ1YxvHxvj2gNhUT+8x26lst49PLCC+uIU/LrL5XQ4K7omXYdRyAXd7KuOF8Yho6OVDdbo EnWmvBAsXvzl9z/08GeeSjUp/FggxKJ5OZ/jc1kAzDQheh8Yu/TBGVLkyCPIN+Ae7zgJzomE sZAYRap44uFdSasdugjjq2/oskIgNrJIclwnF6q7t2LXsLGd3bhu//6AQ1Mfw3zoc7/v0BPT 9xp4TsQBpVUUiUC0lfHk6iB6oUIIS258OziJ2CErQczKYQmpiT8XIiY+KSo0rxlPkqVsOtr7 dhAUa5Col6sFhA1DJNccYvOv8uuNDWmBABknwOhf1xA1BuF6pjvIdxYEkvdZQREQ5Rp8byCk azY+rdQsB2+Nwoc4uP863zBElSMpm3AoW4lva+WWx/CAuAeYMwuif0r259PdSUQkjNoVQU1H zxr6mKVpwwB2xFHK/M3eG4QfkCQUi8q1EzPGvkM3wV2EKHJXDLXA/k7KYr8RTW5Pf29JepsU 8tvRWlLB99Bz8PV0AWkS+vLCaPS8jLtN9yCM1TrfTAQ6IDaFjVklOsEtYrzvKbkCI31ES1De tI2aKVPNSeSjYGUeAiK91HIyvayILZrQfuvf8aIcO+WCIMDSMPsQ8hxN8OwNJpCMIDh5zBVR HXZ0j6jnt3qlkBHFmYwzIicoLkL2lLwJllLBaFpwRAD3DNchC4FmIPDhBMnFtdbeMXMCvjs8 NFtgiMqdJFHBsM4ZqLEbDh/G2WsR155X84EhdXQSUItzCQ9mTyT8Y/YAyuKXkMPFgamyMqmo MSMKUQ083aWmZKnTXlKRsXs18a/SwebZpIxRyvbi1CfsYgTUXc3Fybvstvcg3t8Df4NZX2cz xktBpvY9IgOcaQjxjSuTmSJ5RY2VKhzdKhVNBiv43W5NoJVpvSHc67bWIT+zzTdWyZNwdpKW JicB9MRNHDH3FiBOlTcZ4jeNZa1MgS6hkddUJG2rlwdJ6V2JtZ7e5hB3a6EQX42X6vQ/Ckpe HRejRcFP/FqCivkFK6wfAA51qhvU81Vp7I1vMS6usG8OtTaoUsYCSfaUcnExFzlQlj3NzFQy c3PequgJXxUGDJec1zCbjqaVIVv09i5Ge+Nzp6ocYjVmmsYHYpKULqGx3TTnFAi9/tkYpUsM LJ5t1NtFTmkLmMcyM0H8YMNpMGjAGPg77aiRFcBi6jItaSrn5VpaMSw6gSmDSzhcrmY0e5aa 7rTDAMKY2JLQcuelTWGLPUFch7qKusF9qEPxnS5K6j4ka7aEFwLuQSQGqpLkpEsrQH0sxP2t UdcZwTnv4ONhLufcUCYnMLAZInMfkZms7/3pst8KkeuP7t+MyWgVayhVdDvjWD0CO1NZy/2o sZjtQEqbbGZbwLSzmnlOSTBy20DG5TwfsTWc9k+j0osC32V3uT7QxkufwPelc8rex5fRMU/1 al5t9m3Gi7J6UDI0JXPRZ1GwifjNyPCnJMGfcGqHEd1J1a+VNQSgbIuhKvOi4eJdWxQzqNXS tHGkuViHGKX+LbncBwXJrU1aJNE1DBsmMRMygK+KK4JBOw412ZZQdqgCIwkrQ5CjSUCw9gUW gcFCg9hwms8IgoNM+aX4BkkPzDf8GLlsMMnHXBTm7eqaFbAny+XqipNZELo5fhQ6hlCSwrlT MPoPwn971Hl6VByB3+8avnNdYSL47JAr1uO0NQ9lqO5lCm/4yPHJ6E66ReK4+aco5DExjsa/ hjHg1wR8yHkKYUfqZxp5O1sQs9tk6EFDbcLbLuVtjl1NrUY2Q0j1V0ksjwNao0HjqTcMHTdO TSBqFU5D+k4aX+m01O3Vq1eeFUBfUVCjCS83RXXHPBFLKYz05GUqcR0VnUPV8YDd9XvItLeu 6ZAKUww7RkGs7uWRrS8nUya+eQEZsmjhAU/3Iu8sX1bWtf3e8XJvdXoDfhqgeIKpRsDh9RZl WXWD3xJsk9W4AE8SP/8sf6S3bsTPBmW5DgRed3Y2imlhKzS4wBNM/eWEAwiMuKOnduTV28hO /WodA6+RK7CRhf6KpUkmhdZcUCPLTn0zEf+cCBaN4KVyPPcg2vJCnxbsYXk5zjedg+qma8DK UhoaXnVzVN1cLIuR/Z2LZcC0Ek8LmGZBQmg0szBhNBjddQsaRBUvdBMj6R3Y/AXZ1FS4SDcM 8AWMIQ1XTIhlMS4dqgBYVz022B7YNaPOBHhUzjfwqByM2jHXxk5aT6MXDQFrSEzlNGxYy8eY ph6/QZMisF3BRz93ak7zzLdso0d3pBy1zRbquddusedZ7bTRrVWA8FDki5nQ5xKHoIeJAPRU LW9idSJ1VOQ2aXVcxuQU7ezt6PWAj1/dFnd/+UXFoonaitipfU6jeptd1kPi7qLt6NfdhQaE 9txI4i17EFQxidfleinIaeV09iUUlRpPl/TZpFxQDQ41xJzA1dwzJAw3la2esb+mFMN9rHbB vV5yiapHBi7r9IuiT2NavxJz18Rj9iOoUKmSvJhO13iKisWiXEiWqHebnrAk5MaPILrCJ9OG jekAPoPB4YMjNnoAdcJeRIEm/bOn539/8u2gxqiaCvoP4550R5A4EH/MuD2u+oOYsFxPq9FE 2R+NhX0EJS5t1OcjQV9SNv87cYRoMpDUtUJZEM04IMcNyUmEtVxYx+HCKaZihYg9FIq3PYz+ 2Rc/vPiRK4aww0ePIjrx/42LNFgi1xZ7FH2NLoxty08EyDaHpyVyM7KtMgnGAcACjzFGS6+7 1JLU5axcLw29MLj4q+hfZpAX35LX9CZZoFM+LY/+I+pr78A25ma0kUaFkZw2keWjO0AT5NtV vsi1klr4uq7WfVGDNXvirDOxYvmOkoA7Zsk99c3X33PJBRFjUD9PZGIOC8+fvfj+y0HduKRw mjEt3Wg5JGixW4aLtddacaFZazRzFCQtnGWDmISa4Z6JpUtPt4XOaCbh4ZEf2tRtN69Yb22Y XWYJBgsXCBHbCbTU+PfsugBedlVCRVeaK5p/kFz1+1weRt/RDVUAJn8zEfiUXYojOQR5Q4pF Gdop7HN8xiFOqjnIA4rRrXtbdEkBluKcY398DqkO0UHIkMumc+7Kg0nfrkBZUlIrSvcwXMly ZGe6n7JJQRz3PvKCfSC+uaM9Y7VcwUMfRQUiyj6ykX1x1Nt5hitjGd0KavrLacCmrCFlJ3Z4 3sgNx8VlvppVhtU5SEYG0YgJu0PyAm9N21zEUCwxYmXH3qvPecRHIG0pMpv8C1hatc5jlgAA --------------260F86CEEBFC71FB4A0D3738-- From owner-lockmeter@oss.sgi.com Wed Feb 16 11:11:22 2000 Received: by oss.sgi.com id ; Wed, 16 Feb 2000 11:11:12 -0800 Received: from deliverator.sgi.com ([204.94.214.10]:26891 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Wed, 16 Feb 2000 11:11:06 -0800 Received: from nodin.corp.sgi.com (fddi-nodin.corp.sgi.com [198.29.75.193]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id LAA27906 for ; Wed, 16 Feb 2000 11:06:34 -0800 (PST) mail_from (hawkes@babylon.engr.sgi.com) Received: from babylon.engr.sgi.com (babylon.engr.sgi.com [163.154.10.144]) by nodin.corp.sgi.com (980427.SGI.8.8.8/980728.SGI.AUTOCF) via ESMTP id LAA20516 for ; Wed, 16 Feb 2000 11:10:34 -0800 (PST) Received: (from hawkes@localhost) by babylon.engr.sgi.com (980427.SGI.8.8.8/970903.SGI.AUTOCF) id LAA75136 for lockmeter@oss.sgi.com; Wed, 16 Feb 2000 11:11:56 -0800 (PST) Date: Wed, 16 Feb 2000 11:11:56 -0800 (PST) From: hawkes@babylon.engr.sgi.com (John Hawkes) Message-Id: <200002161911.LAA75136@babylon.engr.sgi.com> To: lockmeter@oss.sgi.com Subject: lockmeter patch available for 2.2.14 Sender: owner-lockmeter@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;lockmeter-outgoing FYI, http://oss.sgi.com/projects/lockmeter now has the lockmeter patch against 2.2.14. The only real difference between the 2.2.14 patch and the previous 2.2.13 patch is that this patch fixes the /proc/lockmeter return codes (seen by the lockstat command) and is now compatible with the latest lockstat command and the latest patch against 2.3.42. John Hawkes hawkes@engr.sgi.com