Date: Thu, 09 Mar 2017 14:52:09 +0100 From: Alexandre Martins <alexandre.martins@stormshield.eu> To: Konstantin Belousov <kostikbel@gmail.com> Cc: freebsd-current <freebsd-current@freebsd.org> Subject: Re: smp_rendezvous_action: Are atomics correctly used ? Message-ID: <3034263.GEZH9i4V44@pc-alex> In-Reply-To: <20170309130754.GX30979@kib.kiev.ua> References: <2092905.6A8RAGlt18@pc-alex> <4355496.HzrDjoMEL6@pc-alex> <20170309130754.GX30979@kib.kiev.ua>
index | next in thread | previous in thread | raw e-mail
[-- Attachment #1 --] Le jeudi 9 mars 2017, 15:07:54 Konstantin Belousov a crit : > On Thu, Mar 09, 2017 at 10:59:27AM +0100, Alexandre Martins wrote: > > I have the save question for the cpu_ipi_pending here: > > > > https://svnweb.freebsd.org/base/head/sys/x86/x86/mp_x86.c?view=annotate#l1 > > 080> > > Le jeudi 9 mars 2017, 10:43:14 Alexandre Martins a ?crit : > > > Hello, > > > > > > I'm curently reading the code of the function smp_rendezvous_action, in > > > kern/subr_smp.c file. In that function, i see that the variable > > > smp_rv_waiters is read in some while() loop in a non-atomic way. > > > > > > https://svnweb.freebsd.org/base/head/sys/kern/subr_smp.c?view=annotate#l > > > 412 > > > https://svnweb.freebsd.org/base/head/sys/kern/subr_smp.c?view=annotate#l > > > 458 > > > https://svnweb.freebsd.org/base/head/sys/kern/subr_smp.c?view=annotate#l > > > 472 > > > > > > I suspect one of my freeze to be due by that. > > You should provide either evidence or, at least, some reasoning supporting > your claims. I curently have a software watchdog that triger and does a coredump. In the coredumps, I always see a CPU trying to write-lock a "rm lock". Every time, that CPU is spinning into the smp_rendezvous_action, in the first while loop) while the others are into the idle threads. The fact is that freeze is not clear and I start to search "exotic" causes to explain it. > > > > Should this function be patched to use > > > "atomic_load_acq_int(&smp_rv_waiters[])" ? > > There too. > > As a side note, any read or write of the naturally aligned integer > types with size less or equal than the machine word, on all supported > architectures, are atomic. The meaning of the word atomic there is > that when reading, you always get a complete value that was written by > a writer into this location, not some out of thin air value. Similarly, > when writing, you are guaranteed that any observer of the write will see > the value you have wrote. > > The guarantees above hold both for C-level code and for the assembler > accesses. > > atomic_load_acq() provides additional guarantees which do not affect the > value read from the variable itself, but establish the ordering on the > visibility of the related operations. OK, I got it. Thank you ! -- Alexandre Martins STORMSHIELD [-- Attachment #2 --] 0 *H 010 `He 0 *H 00n 0 *H 0H1 0 UFR10U STORMSHIELD1#0!U Stormshield Root Authority0 140904150710Z 240901150710Z0I1 0 UFR10U STORMSHIELD1$0"U Stormshield Users Authority0"0 *H 0 X6[t. D Uge0-2;h@eȻClΫp B# M,FY=.{ya{2πߢ 7 <d~O;ޅԋ&C 8ُ6@CXX>|abqenmI, O&'۰@%OhW& {52D%_8#f]G0ct y \ v0t0 U ml||uu4[ׁ0 U #0BgDaP0U 00U 0 `HB0 *H N9-؞>m-K!M-7zD6IZʾr>q? u ?xr6_'NջZ7]V\5Y&) )m@~^Qdp/H3ͦ` o⬋U,z,0Bdp!C2K8.r>0-1!C%3U2ϠMg~3 ͟wͲAZ&vF/BA$_ڀV! s7)=-nP> qH~g/Rs,PAJVm#QXJo */V|s Ŀj~y߽j 3]v3ƌqe`Y|HB3lC!^J2$4A`Z+( 7b{e˄wʈ<L=6[3`>IQF aiܝ ?:_yԊ / `0W0? (˻0 *H 0I1 0 UFR10U STORMSHIELD1$0"U Stormshield Users Authority0 160901151108Z 170901151108Z0p1 0 UFR10U STORMSHIELD10U Alexandre MARTINS1/0- *H alexandre.martins@stormshield.eu0"0 *H 0 ~ {x~;#3BgXC[rSVuv#>aL0w}""v`B)Db8qHkH\d qB6rd|:%Ze[wi3)$! hXε< صVO}# EjJe gk0%L퉬Q'b}3"*(_ T- w?gR H,\aWO4Ǘ Y3TR(Wn7Sv1n8xIan 00 U D0IM_W]A$v#<0 U #0ml||uu4[ׁ0 U 0 0U 0 `HB0 U %0++0JU C0A0?=;9https://pki.netasq.com/auth/certificaterevocationlist.crl0U 00U 0+U $0" alexandre.martins@stormshield.eu0 *H 5hhN ̛U92+ E jr@|f{`J2X!ȳ.ڤ|^*$"vlHtSeeAܟ-bC$*)Ж WJL >5@|N%·%{i-4akG앏]%.ݤp]7B/ *1WE@ zZoq@fk0NpHObD $9jXḓSk~LKX100R0I1 0 UFR10U STORMSHIELD1$0"U Stormshield Users Authority (˻0 `He 0 *H 1 *H 0 *H 1 170309135209Z0( *H 100 `He0 *H 0/ *H 1" d c< ͨʐu(&k^ZJ%0 *H dV`Jc.2i@@w65e ?" XuNZi&{vg酓礗,\Q\:1ذj @i>X(+IJHP^]% Xg3ēBrUb?!: c;A䝅T ŻoN0LG}T cC!Հ Z{1ԡNtPMO? ЂwuؔUHk';$ߣ {AmQވypeF5Î/gThelp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3034263.GEZH9i4V44>