NetBSD Problem Report #54210

From Frank.Kardel@Acrys.com  Thu May 16 14:39:28 2019
Return-Path: <Frank.Kardel@Acrys.com>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits))
	(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
	by mollari.NetBSD.org (Postfix) with ESMTPS id D162D7A149
	for <gnats-bugs@gnats.NetBSD.org>; Thu, 16 May 2019 14:39:28 +0000 (UTC)
Message-Id: <20190516133413.C618569E23F@sf2.hw.abs.acrys.com>
Date: Thu, 16 May 2019 13:34:13 +0000 (UTC)
From: kardel@netbsd.org
Reply-To: kardel@netbsd.org
To: gnats-bugs@NetBSD.org
Subject: NetBSD-8 processes presumably not exiting
X-Send-Pr-Version: 3.95

>Number:         54210
>Category:       kern
>Synopsis:       NetBSD-8 processes presumably not exiting
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    ad
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu May 16 14:40:00 +0000 2019
>Closed-Date:    Mon Apr 20 20:13:49 +0000 2020
>Last-Modified:  Mon Apr 20 20:13:49 +0000 2020
>Originator:     Frank Kardel
>Release:        NetBSD 8.0_STABLE
>Organization:
>Environment:
System: NetBSD sf2 8.0_STABLE NetBSD 8.0_STABLE (GENERIC) #4: Wed Apr 24 15:54:41 CEST 2019 kardel@pip:/src/NetBSD/n8/src/obj.amd64/sys/arch/amd64/compile/GENERIC amd64
Architecture: x86_64
Machine: amd64
>Description:
	On a large memory (382Gb) machine / 16 CPU a postgresql 10 parallel restore (8 threads) does not terminate 
	2h after the work has finished.
	The stacktrace dump shows several processes in exit(2) stuck in the VM system.
	After "reboot"ing the system the DB shows all work committed - only the final exit(2) phase 
	of the processes seems to be an issue. The DB processes also have access to shared memory.
	We try to repeat the scenario to gather more trace information

begin 644 trace.txt.gz
M'XL(")YEW5P``W1R86-E+G1X=`#M75N/VSBR?O>OZ,==Y&S`^V5?#F9W@\5B
M!S.#S.#@X+P8%$5FC'3;7E\RR;\_I"QW2RPJD9/N5HM.'H)6B998Q>+'JF*Q
M]/>=V?]^\\'M]JO-^D:]1LM??_OA;S^^^:^;U9UYYS)W7B]^/AZVQ\.-WVWN
M;LS-[KA>K];O;O:?]@=W=[/:WQS7.W>[,M6M>[WX]=-=M;F]66]"^\UQ78\@
M'';&NK_>;%?U#:&<JYO;\!>^,8<;]%%RHY"0Q"&$%N_<VN^7MQO[_D]_CK<?
MKE^ACTQV'Z2X(-WG^/C/U;6MK$*U4HOEO]^\_>G-C\N??_EM^=,/;__^TYO?
M3@]%W<=@H92"SS'2T=`E&?KTF><LES_^\/:?;Y:__!#^^_5?__?F=(,;0ZWB
M=?\]',ON>WJ]D$3@M!,*X<IH*H45:+&_=6[[GV7U()HN)0A'RX7]L/S#K`[+
M_>K=J4F'$%H8O*@WRS"F#7%5G]KT2%'(#'6;@4:AB?*+>+5LKAE';7^ZI/@@
M$EM9<WM[?S]>A#O8V<5?_O*7,^7F3^OC7>5V-XSI/]^$&PO):L4,H@X)\]>>
MF*1$0V)RSH\3T_NM.>[=Z>[I[T"OQ2)(:7_<;]VZQNU/'PBA!:$MUQTRZ[+>
MI8?VM+Z4?Z)9RS\2-7(D\&_Z_&/)"<\*`#ONKD]/A.><$*@GF'.<EQ.UMB1%
M\9PPY3**PC33&0$86PE+:C).`'MWNXQ#:LVZ;?)`B"/GJ\5V<WMK-W=WF_7K
MK=D=7M-30T".ZM#PN8RW'J01K^++JHME@,Y*H)FUM:6.J%0&'+&L$MA:.6/'
MR>!PW*WWA]6MZS9+B''0F5_L_EA^</:PV2W=^N!VIZ8),7(4I/_A;GEGMIV%
MKD,(3019'%N*^QA9:N=70HPM<=-RM7EHL-H$N@W3?+/]M%HO/]SMMT$F[4SO
MT>*@W+?;[C:VUR@2HGKR^Q;[_6$7[()]O]F9&MKRTS`>;I?OW>[TA*79G7^0
MO17?(.]_5J]"QP[V]]X/SL30M#+GN=7<ZLVJAA)GJ+I<ETBK2U)S5VE$N4QT
M22&97:`1Y]P$@^/+NH21B,A[6-VY&L!OCQHQV"S>NP]!8UJ$:2^BL(1N17"B
M]6'U3(L-Z<6X2GDK!HPLL88ZAA(Q4"QRL!+6G\H13U&CC-X<;P_+553WM;E]
M4,P^O5&MQJ+;MI,J_-4L+*KI7;R\[YIH.U8[7N%>EPB2)&<3F!HA)HI".N&5
MUI)!I*-\8%"\,&Z,41!U<VBU6YOUIFG>JN+#=809UBKC/;6OCQUR%./%,Y-1
M=#:)/%(BK'3$F63\.1HPB6I9UOA+P<+D0'#\A="M`M!6"%ACIWBVC12IG&R8
MNGBDEMS^L5T&/ML&YZMP,]I4S9`OST31TX0..2X4EQM]4K52(,(9CW3X-R`%
MR5.,UIS4XXWCT1"]WSR8OJ>_&TO@;!>;8&MN^Y;QB139/)F]O3:]!DQ<*J'[
MJ7(2$'6<#`F(`0%I5]PB=A9#NH@]B(&F,T'3\M;R+XN!0#&,`X0P76:!!\!F
M"^9_40PB:)06-8+!Z.Q[L=YQ412#"C!(75$,2F"=5>,8G(O5(8!=A=4X'9T-
MAZE=1;D9AZ.SX9"E8ZA181P"FR>8_V5Q2%(M%:@L_T8`>T;[PL8P-6B8*PQ+
M>6+1U`C)PCA4"=(0Y,O"4I[:-('!LI"&@_50F2OTD#E8-)$L;*!!#$"5ILH@
M/*Q$41X61Y#!LO"6:<AA64K*TC43*UZ69<?2-9.:PJP"EL8!J*9E67:,IUHJ
M=6$<IG8/M;8PI$E,&L05*FP,^R:-\;5RA6$IR'C@=6%:"N(`]<B=F[EP2%.;
MAN#"(HXTM6FHJ<(O'G\WOMXLW^_=73\=]9X2VLC+$RTY27S$>BCU@LH44+4O
M:@^.0KNFK$U&"LP:7Q>U34S[5DVS(I:U7E`0J%%E[1-3&*<I;!+",`TN:Q+"
M,`TN2D4)C-+@HE24@""-U$5%$@G(U9"ZJ#E(0*J&U&6I*#!E""E+1<'&%"%E
MH2BP9`@I2T6!)8-M62`#+!ELRU)18,G@D4'2F3"(@26#55$HBH$E@U59(P@L
M&23+&D%@R2!95O@0IZ8,\B.33N<RA.`PCL=EZ2C8:O)EN?0XS3@-#)8%HR#A
MU/FB3!D,#M!4(R.C,V%0`_Y&9L[,A#\%^.-%00S87*I046N$2/G3905%TY,S
MK"X+7T`XIAYY(F$F_(%HC"\LW-3CS_A:^L)J)YPW\(=K)Z1#7,814RP0H]G*
M-<P[44NTV,:*0CMWM_G@EMN#.PT%;T<H>^]54QBD<P\TCETVK*D&<URG[5)J
M;(R;.D9-5:*EW[F'E@^DT"Q(P7U<G8\)-'_&'^M3%9)X_2#<>/7J:ZIKX5ZU
M#>\X2RJ.$$%`Y<'3JF69J_DXE;F@KM3=\>`^9DI+07KDEYF%B;*-)1O]J>'#
M=63;=X:E=@=SKK:44E\U%1='C(I_":,B,<L6^XJ#8L9"V5,.RK%3XBO>V'UZ
M$'N?'J507S!WT(SF#L=Y,`JC)!Q"WZ44I<0%RI:YC&ZI<`Q=Y00GBDA0J/7L
MS+KO,_R%Z"Y&J%_GMCM,XOOJ.-'D87J@P+#2>/3R.`]+W^!8Q2E3#Y03AMKL
M@39>%,ND$3S4*-T\(GA601=>U]*XC$?083'UVS6=U0[G"!9!VK*FLPI]CF`Q
MS?9!7,UJ&W<,BR!;9&RYA+FPB%'*HL:%*2JH/$G=O`Z9CV`QW:VFIA)75'KR
M)"%8>K(CH73)L=Z*XDI*G.4`"B@_R"$-*=.*7*,<0&&X>FQ=AMF``J@,5R-6
M&.ZEI>$,06Y6QXE'L`CVN/#(-);YL`@PR8Y,9)D/B\#*HJJT400N/A^9K3,?
M%M/48STV,W<V+'*8?#PR+7`^+*;IQW9F5?Y&L`BL&SZO4BDC6,P8].45D_^R
M(0N*Q`4KK["5!52)"]YM:9B4?B_+FUF=VAG!(5P[2S,/DD)Q,9U)V,)8!`7C
MS;SJ.(]@,5T[K1]9ZV`^+*:1`>M'%HV9#XMI9,#ZD6>MY\-B:@$I65I\)RT7
M%U@LS>%BJ7%#4'%PDT8&D#>E+?WI21XJYU77\,LL@I)Q2L[KS.<(%D'-."O<
M[&K&G7W&M&9<A\^^B1-<QGE5.QHQDN"`LA]9HW(V'*8&CJ:B+(\QK1L7M;2P
MQ1\4CM.XL/P24#E.N5E5K1K!86K=:#2KS_R-X+!OW,2XAAP7UY@+A[!XG!\9
M+Y\-AV#3HQYI@\^&PS1NH]&LJ@6,X!`DLIF1GVF8#8<@C\V._(C:;#A,@S8U
MFM5YSS&IS\"FF57EG#%IP<#TK@L;0[`=5<VJM,P(#M/M*.M-65@*RLAI.O+K
M8;/A$'[J;U;E5T9P"&P:.JO:\",X!#:-'5F/<S9./B@E5R-6UG*1UI*+"1J%
M04W?J(FK16%AC+2:G%&J-"U-`S7!JBD,3%.CAO#"S+;4IJ%F5@6?OLP@"-.(
M696G'G%4+8VTJ<+B4*E!HT1IF7SI"%:S^DS#B,W\9*VO)"TKW)U&:*P?>6IJ
M+@RFMHQUI+"Z<N=-_N%J$TD(QSM5Q/ZP0C1;CB=^[V=<_O^%N1I3IO]K9AW-
M?"*>#U15H5(Y._[8-V"_RSD^'^J.Q%5]:M,CQ=G`4+<9:!1G@&_E%*]97TPM
M*3[H<E5AY[F`.7/.4H=$4GV&<C50]DO[X(4_&AX$2G?>D_MV/6KD156M,,)-
M9Q.E.=-"0^$O%@>6K3@JJ@0FU.%$:8A0-%NKSMA:%0:/9QD`>$1(Y>IN-A\E
M%-9/J1+;G$YL.THA+Y\C5+0"\5QH'0625&A2/%_SS5B$R?5!2868H@)"B91H
MH+J815>(N$(:3C)B4DKF$0;7879=G9@J3;`G4$Q8,)E?P(,VD8)1B#A>5PZB
M$,$H7V316%Y=(0PI:;$R&8M&,3I05;GV5SC!&&:!_8R<L.1Y=1+V"L542Z>C
MLPAQ".<-9&.9-07C4&".X"P.L1P,-1F/[/KTIK*<HQP,B7R0(<*UN<+Y%>":
M!N<*R@E1FB_9'N#:7)^<+.9*YN1$L\9UW()T):,0L<XAB$)R(,)%*W>%OH;G
MM60Y$$(!AO)J(_45VHPB+%"Y*!AFC(FA:.$5@I#V+*A'SF9D0Q7$B21EA<<T
ML[:V,#Q&!%<#Q>ZUOD171D?7#X=/S0-.;<Y745)UN+=SIC[?B7\'.M%BL3WL
M'^ZT%Y'OFB]L[3XL'V[>7YYZMCU\\OME][?GZW"_4HO_^?F7Y=LW/_SC=/=\
M%0="+3ZL.S]L+^)38_%3O[IU#S<?KB./JAG"A[OGJW"/7QSEI><!K$X!&)(6
M/,6,#PU@?86@2!$-#A^<[)JQ_%QG5QFH$E@CFUL[I$(#!JQ35ZA.M4&:Y1QI
M3/(!A_A%R"NTV1CC&F?D1!`;<(BXJDH..-1(!",B$W!`&.>_N4/%-6XLU,*Z
MG(?(U,#7Y+B^QK@G0957F8V%&)C)KVN4FRM4IV#@UCZG3@(-?IRP,%.?J+BE
ME#'U.>:Y)2N@<%W5!4,Q5](:!:&8"I+](F>8.O0*31W*M*&9K95L^D3SE:UK
MC"54#.<"OQ0/F#E47N.JKD2M<B$\3+`>^GHF*GD?7&!CJDSD%VM)LXD!86$2
M5Z@XVE<FMW%)Z)#9S&I_A6!MB6$D-\$HE@-X;>45XG6P[9C+^>]!)GDQ\8O<
M]^]ASZ<->U;8$&5AV),CD=OB:/;MU??Q>S'CAY&NZLSX84SXD'^/OT_`ES.`
MO+):F-P`ZJ&/TROT?>/HY0R@T48:ED/0(4N\0E<88)-<L)SAJ1`?,!3L16I>
MAIA$8+[.YE^BX,D,;4->H3K5=>UI)DZ`A1(#V[5&7J$?8PP*KC&4$T<#NVOJ
M&L^[6>M]+IRB!,DF^@0,O\:S/,9P5+/<)HEB`R!>N>L#<4E5A;.;VD3QO#Y1
M<86+G505\R27O:JE&`BR&"^*/6;*N'+<9HZ92CW@R@GAW,B8[ONM.>[=Z>[I
M[T"OQ2(HS?ZXW[IUW1[([A"BC4_/?#^068_W#CVTIY>7V=<LX1^9](P.(4.;
MB^(*;1]786US\(+DD,NAKS)G!E5"YO)2B4`#-J)RI%AXX5SZN+4&X`5+F0_]
M<\L*V[L_RP"<8J<LZX8V13Y*PMB:2%D)B+&8,9GS+V/`U]OKPPYGM$:Y_$TJ
M62XL%^0D*E*0HF`;5#^S&&M$<FM,\VE&51#_M6)593(3A7`QD.7CKC&N)ZJJ
MIKE%5HGL1JGQE7/$QU4E#-8?JX/]_;7=K/>'[6ZS?2T?AAW<BNR8Q6H?;JY-
MV]/S550.<ZI>&Y1IMW.M-#J$R`W'0XP24F4912V;".G*!S:I2VV)?,Y7C2@7
MY7T\V4H9=\73ZDF,J>QN88WBD;SO>Q4O9:^":E/%\AGI7@43)'M*NPX.]Q7&
MN;!2=>Z,"\L7&0EBTJ@D(Q%9Q>J<(\ZS-7N::7Z%2Y]M%K+,TL=1UD2JD:I*
M4A,O?/"6H9HPFJU9U/!_A5%.3K&M,N$:JF76DXAH(@I2$RNDXQF7D^7WH!K^
M2_*DD%,:LPR:#GB2<9J4-/ZDYD1FQI_08:/Q>X;2RS$:N6;$9U+,&,V7]JD1
M8N/JQL[)]<$5LU9!UX<&U1XPG=4E-M%W+7[B-*VJXG7F?#\5.#M^UGM;;D5;
M*0RI,Q5MJ<ANO#9?"BYL*X"BVO+,5@`A/+LL6Z^%LT_PZ?*OFM"?G[5VW;FU
M;NG4?N%78;FWW6EWOHSSKAZ>Z1R]J)DN?"5\=$O`>J6'8AP7I3Q-.:YEC%#P
M""J,<CG/,C_UXA"]G)3G:QBB,(FH8)E(H<R>*FC"W2\GT'L-(R29P,[G"A8-
M[+L$6T8Z-V:(HED^Y%>NS7K3-&^M\(?K<-^RUG*YI_:-EPXYCJ"ZW!H_;\=(
M(Z6,Y4;3[1B4/<48]9.:XKYEH2W'F6]9!"$,[4G1\O:D*/'291RS?%V<!JG*
M\TZ1C0<QH1"PS";FQLUJ7FZFDZV(BGOWJ7>#:;8F>@.-Q:&#H1*S##I@I0<V
ML1`O*>R(+&959G<"2YP-.]I*8E>/Y'_$[L3N@SW<+E>;\']K$W0HD;=VYZ'3
MXOXR,HC-I;SS,^L&5=([[YQ-(^ZG=8&(]GMEM(H'UR0*0+#\]YNW/[WY<?GS
M+[\M?_KA[=]_>O/;Z;TH^P0`)W3<9S%?R#>&C<6:,?@M+#94P2>H45GACY,$
M8/@C>\HLEC#2;I0`X@#/0P)#)5MU'AZ"!,:IP/PEP//V8Q1!85^U&I(`RR;2
M-Q(8%P8M00D&+,?@>UV)"/)EOHVMI:97(H*!/;XH@NI:1$"S\13K+;:C7*B1
M(GA9/M19&JD/Q?(!P"`-5)J!(!AC\2P^4`@Y8"'8D4;B_$7`LK5)&A$49B$,
M*D&V+&PC@=(LA$$E&(HM5?A:H("Q`4NYPM<#!=F(>Q3!."@H0`3Y1+[XF7M1
MUVAA[LQVN7/^],;S5?/*Q?'#W3(0W/JPM+>;=1M32ZFQXTW;_3:\=>DWY^A)
MCQ2?6-7-([>[C>TW>R#%9O4B_M4&Z9H_8TC*->=C.K\[7\6;%X=;2$]LE55I
M`FBV/+"Q1GI5KF'EE*;Q0R/@&.Z`F5D).RXV/:?]"J&$%)F#1&0@\J!]<+U+
MW:ZH).?>9;8K\K-#U5=X'I<Y(V2F+M9I[2&8]J54(UDY?YF4>A(Z`>X?`?K^
M<W1'MXQ_N=VI24J-;7&N4VF4O)+QG.S(J;SUR^UQ]\XM#[]WTB_[Q*AX,O?F
MS-X&\>,WN`;$L?/+MV95_^OGWSI=2HBQ]Y+F^H0`L/E'&*)__/#/-Q^=/1Y6
MFW6W6Y`>>\95IF=(PYZ9B:6%%.R3>!G2DJ!GR$XM+0$7S1>B6QR.(YD:F!`#
MG<+6ZV_M5#`RMW6PLS=AN=V9U;J'7`,W3Q9YIHL4=E%-+C<`Z`(SS[XQI6IO
M?W=U6$W7MOWI_76S,O)<1Z"1B)&3XZ1S=^ADU+47X8ZG)\?!O'.;XZ'C-IP(
M<:DA#/8%ZQ2IC/;5A2/U&0.E<K>W/47J$%XU^7.9+J409:2[%*(^TR5SK%>;
MY?;6?.IU#)!C]S*CAW6*5D%B\L+I]\7N[9S-].Z!&@UXD^M<B@U&,__M\+ZW
M^]5VM0SV]?;61=SL=6[H;C-S,O"`=0H/P6>T?LP$2+TDZ"%%*<K%<5\M&U^G
MCV()-3*7'>(4*8Q"["7U+^-53`VO6*<68N@4^_9.'=?;OHSNKZ-/0C/V!%:I
M11AZ@J86CP)`JZR=O%,`:I6=7)$4`%AE'T&1OK%3J248.C6]3@&T5]7T.@7`
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MURF2=DI[6[VX[`Z2KCN<.SVYZ-)U)W2*3]ZI=-WA1+'1*<:KE5WN[ZKC?KE:
M'W:],<O>"C]C.M>+=*%A6M;>CA.-.9CXFEZN4I<46E59UM.%)+R4/_E+TX6"
M:?'TG*8+07CIDW.*4Z!GFC\YIS@%\O#2I^<TQ42FV=-SFD)<>.G3<YI"&%.5
M&965]OD$JWVU/)C]>Y")UB&&EC:7B(93``M=$A-W":"9DM63#PU`,R7ID[\4
MH)D23\\I0#,EGIQ3<,2#*?[DG((S'.&E3\\I0#-IA:>/F'FZVMZM^K;"`R&T
M(#Z7)0Y.:#")^;?/\X_]).>/G>QF+G+=:!$0Y>X!*!)>QP_5_>]^XP_1##J]
MY/XRFD.^.;<6KMQN=]P>FL-JL4I9>/S==G.[6KO,<34,3C<PB;0S8P:I"I!U
MW*WWA]6MZ[9+B'$@6"XWOGN&P3O-,')(T?AJ=[O>'%;^4_O>\^6KYG!CYD$`
M/**X@CUL['9EM\>EW1_,P2U7=>B3BP)JM7CH=GA17>5^/?B[TR^:)]QN-NV)
MAOO+*"V,%W?U<K?9')9[=[";=7L&.R'FV`,P%6;1LRNLUH/ZFIR[B!_#=LV!
M^TO5]:0)QELKG116913V?*`"+5#NKG@<E=+`/!'QX[/H<R-L[5<.,#CG$%Z&
MGW]\Z?#X`HM`5/Q)X`@<!F#"\F=!HV[&_[=H#LC7;V15"A:!Q'\FC']V555R
M4%5!OCT3FCZ)JH(D>A:K"#R'JG8SY;])5<'2'V7U>9#[!M6!(*+JYU<=/*PZ
M<)F5^$E4!V2`!U'@T:IS.I1Y,.]=F/?GL&:'%)<B?7[4SMUM/KC>LTZD"!+^
M)-?W[KA=WM=[Z9%>Q>+`B[^9^L8'AMS-=M.P^M<'[2.5-U9PEA%H-Z?\6_04
MI($W`U,*I()\<B:"Y_O<\T(.>R,@C;NI)60NGQ?_?6I''`GL!7N*$8^P,)]5
M+X:<MXJXS(%VD,G-!/<^JXB/Y&B`C&@FF'\JJQ!D.H>7Z6?7"S%L]8,$8R9(
M_21X><X:QA31LQ;=PT\5],B@7/\>R1<`F;@-GZ7`#TCI#=S)YU>S8><#9-(R
M@:JG43/\56KV2(X#2#EM^"Q%S4#N:N#N^8-N?-AQ`"FC+&[T/H6:@3Q0QOWS
M1-Q`LF=X<^.;WBW?N5:\S9_1[`LF9&6V\?L7IJY;YKN4.`#U8F?^6*[6VW.5
MCOO+V%N9V5`'J9VLW4\O0\L!5G'W_)$Z/NSC)*F9%>*5$PY?KN7[U;M5\W&4
MU;HM_M&E1"%FQ-/-P3SU(!CRPI$NV+9BJV5,'8!/>"27`N0X,FZ?++X'<A?#
MRYX_OL>&+7R0,LBX>9KX'L@#9+QZGO@>>R0?`.3G-;)ZJJ`)2+QC7#]_O(T.
M.P%)OEME<.5(S;X^]%\)8JARN=`__2HW@#Z2&T#ALJV^%"W[:LB@<$;*YP^6
MT6&KG/97NK"0Q%T+=_FXNX^K<^'7YL\X730;(L<ZGO&Z9;V]>O4UG^?"O9+`
MWG&6*^=)X727>.RGR>#3'LE7(,!7X`(_E2X28+IS_OP!*C)LNG=S-$^=U!(C
M92_6Q9JK,";A`9IIXB3+E7>%*,#\LRQ@Y)&B_C!#DM,G"V+!S$A.'R&GZE+=
M&3:("32(A2L0QV`.);\@S6.E"%-ZN=JN6MVXOPX-5"YS[9'L=)@024Y%H8IP
M%V'F)5-FZCQN`'!,B6\OR!JT]BYFV\><QU,O>FGY`W?C;W.5(8%UQ-34E6QA
MJB:34Y?]AIF<3#Z"?NTWT<WW.Y<,8DINYD8N]0W85$P^0D'I;Q,5<+28'%>F
M_'/U=*VQO_>%U*5$N,]EVX">"#<J:_*STOD`ZR!_2*H?4Y&9:\#XBQD?7_%]
MXOLN)2F<=YOZ>-N748_42#BSPH"%@8GG-RS$H%W105'O:HZE55J&]2]U6YTU
M&MG/;K$;8J07X5\N<@.DP"L4O.]ET/2UNUT>UP]RZ-,"1R[QHATBU'&%LEYT
MZWJ/_3IO:UV![B$5UY(7MGI#-$@<"JQJX<9^0&@(#&(%ZR91(RAWRVB/%$V@
9YGLG\=/,P<+;G(MP=RB+_P?8L5;\B8$!`,9^
`
end
>How-To-Repeat:
	Run NetBSD 8 on a 16 CPU/382 Gb machine loading a sizable DB into a postgresql 10 instance with 8 threads. Storage is NVME.
>Fix:
	n/a

>Release-Note:

>Audit-Trail:
From: "Andrew Doran" <ad@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/54210 CVS commit: src
Date: Fri, 13 Dec 2019 20:10:23 +0000

 Module Name:	src
 Committed By:	ad
 Date:		Fri Dec 13 20:10:22 UTC 2019

 Modified Files:
 	src/external/cddl/osnet/dist/uts/common/fs/zfs: zfs_vnops.c
 	src/sys/miscfs/genfs: genfs_io.c
 	src/sys/nfs: nfs_bio.c
 	src/sys/rump/librump/rumpkern: vm.c
 	src/sys/rump/librump/rumpvfs: vm_vfs.c
 	src/sys/ufs/lfs: lfs_pages.c lfs_vfsops.c ulfs_inode.c
 	src/sys/ufs/ufs: ufs_inode.c
 	src/sys/uvm: uvm.h uvm_amap.c uvm_anon.c uvm_aobj.c uvm_bio.c
 	    uvm_fault.c uvm_init.c uvm_km.c uvm_loan.c uvm_map.c uvm_object.c
 	    uvm_page.c uvm_page.h uvm_pager.c uvm_pdaemon.c uvm_pdpolicy.h
 	    uvm_pdpolicy_clock.c uvm_pdpolicy_clockpro.c uvm_pglist.c
 	    uvm_physseg.c

 Log Message:
 Break the global uvm_pageqlock into a per-page identity lock and a private
 lock for use of the pagedaemon policy code.  Discussed on tech-kern.

 PR kern/54209: NetBSD 8 large memory performance extremely low
 PR kern/54210: NetBSD-8 processes presumably not exiting
 PR kern/54727: writing a large file causes unreasonable system behaviour


 To generate a diff of this commit:
 cvs rdiff -u -r1.53 -r1.54 \
     src/external/cddl/osnet/dist/uts/common/fs/zfs/zfs_vnops.c
 cvs rdiff -u -r1.76 -r1.77 src/sys/miscfs/genfs/genfs_io.c
 cvs rdiff -u -r1.191 -r1.192 src/sys/nfs/nfs_bio.c
 cvs rdiff -u -r1.173 -r1.174 src/sys/rump/librump/rumpkern/vm.c
 cvs rdiff -u -r1.34 -r1.35 src/sys/rump/librump/rumpvfs/vm_vfs.c
 cvs rdiff -u -r1.15 -r1.16 src/sys/ufs/lfs/lfs_pages.c
 cvs rdiff -u -r1.365 -r1.366 src/sys/ufs/lfs/lfs_vfsops.c
 cvs rdiff -u -r1.21 -r1.22 src/sys/ufs/lfs/ulfs_inode.c
 cvs rdiff -u -r1.105 -r1.106 src/sys/ufs/ufs/ufs_inode.c
 cvs rdiff -u -r1.69 -r1.70 src/sys/uvm/uvm.h
 cvs rdiff -u -r1.110 -r1.111 src/sys/uvm/uvm_amap.c
 cvs rdiff -u -r1.68 -r1.69 src/sys/uvm/uvm_anon.c
 cvs rdiff -u -r1.130 -r1.131 src/sys/uvm/uvm_aobj.c
 cvs rdiff -u -r1.100 -r1.101 src/sys/uvm/uvm_bio.c
 cvs rdiff -u -r1.211 -r1.212 src/sys/uvm/uvm_fault.c
 cvs rdiff -u -r1.50 -r1.51 src/sys/uvm/uvm_init.c
 cvs rdiff -u -r1.150 -r1.151 src/sys/uvm/uvm_km.c
 cvs rdiff -u -r1.88 -r1.89 src/sys/uvm/uvm_loan.c
 cvs rdiff -u -r1.366 -r1.367 src/sys/uvm/uvm_map.c
 cvs rdiff -u -r1.15 -r1.16 src/sys/uvm/uvm_object.c
 cvs rdiff -u -r1.200 -r1.201 src/sys/uvm/uvm_page.c
 cvs rdiff -u -r1.84 -r1.85 src/sys/uvm/uvm_page.h
 cvs rdiff -u -r1.113 -r1.114 src/sys/uvm/uvm_pager.c
 cvs rdiff -u -r1.112 -r1.113 src/sys/uvm/uvm_pdaemon.c
 cvs rdiff -u -r1.3 -r1.4 src/sys/uvm/uvm_pdpolicy.h
 cvs rdiff -u -r1.17 -r1.18 src/sys/uvm/uvm_pdpolicy_clock.c \
     src/sys/uvm/uvm_pdpolicy_clockpro.c
 cvs rdiff -u -r1.72 -r1.73 src/sys/uvm/uvm_pglist.c
 cvs rdiff -u -r1.10 -r1.11 src/sys/uvm/uvm_physseg.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

From: "Andrew Doran" <ad@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/54210 CVS commit: src/sys/arch/amd64/amd64
Date: Fri, 13 Dec 2019 20:14:25 +0000

 Module Name:	src
 Committed By:	ad
 Date:		Fri Dec 13 20:14:25 UTC 2019

 Modified Files:
 	src/sys/arch/amd64/amd64: machdep.c

 Log Message:
 Break the global uvm_pageqlock into a per-page identity lock and a private
 lock for use of the pagedaemon policy code.  Discussed on tech-kern.

 PR kern/54209: NetBSD 8 large memory performance extremely low
 PR kern/54210: NetBSD-8 processes presumably not exiting
 PR kern/54727: writing a large file causes unreasonable system behaviour


 To generate a diff of this commit:
 cvs rdiff -u -r1.343 -r1.344 src/sys/arch/amd64/amd64/machdep.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

From: "Andrew Doran" <ad@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/54210 CVS commit: src/sys/uvm
Date: Mon, 30 Dec 2019 18:08:38 +0000

 Module Name:	src
 Committed By:	ad
 Date:		Mon Dec 30 18:08:38 UTC 2019

 Modified Files:
 	src/sys/uvm: uvm_pdaemon.c uvm_pdaemon.h uvm_pdpolicy.h
 	    uvm_pdpolicy_clock.c uvm_pdpolicy_clockpro.c

 Log Message:
 pagedaemon:

 - Use marker pages to keep place in the queue when scanning, rather than
   relying on assumptions.

 - In uvmpdpol_balancequeue(), lock the object once instead of twice.

 - When draining pools, the situation is getting desperate, but try to avoid
   saturating the system with xcall, lock and interrupt activity by sleeping
   for 1 clock tick if being continually awoken and all pools have been
   cycled through at least once.

 - Pause & resume the freelist cache during pool draining.

 PR kern/54209: NetBSD 8 large memory performance extremely low
 PR kern/54210: NetBSD-8 processes presumably not exiting
 PR kern/54727: writing a large file causes unreasonable system behaviour


 To generate a diff of this commit:
 cvs rdiff -u -r1.118 -r1.119 src/sys/uvm/uvm_pdaemon.c
 cvs rdiff -u -r1.17 -r1.18 src/sys/uvm/uvm_pdaemon.h
 cvs rdiff -u -r1.4 -r1.5 src/sys/uvm/uvm_pdpolicy.h
 cvs rdiff -u -r1.23 -r1.24 src/sys/uvm/uvm_pdpolicy_clock.c
 cvs rdiff -u -r1.19 -r1.20 src/sys/uvm/uvm_pdpolicy_clockpro.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

Responsible-Changed-From-To: kern-bug-people->ad
Responsible-Changed-By: ad@NetBSD.org
Responsible-Changed-When: Wed, 26 Feb 2020 21:56:25 +0000
Responsible-Changed-Why:
solved in -current but there is still more to do on postgresql
with huge bufferpool (or whatver the PG guys call it) i see N second stalls on postmaster exit due to pmap work. maybe we can improve.


State-Changed-From-To: open->feedback
State-Changed-By: ad@NetBSD.org
State-Changed-When: Mon, 20 Apr 2020 19:22:59 +0000
State-Changed-Why:
stalling during postmaster exit should be fixed with recent pmap & uvm work.
being realistic about it, this work isn't going to be backported to 9/8/7.
are you happy to close the PR (and use -current/10) or want it kept open?


From: kardel <kardel@kardel.name>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/54210 (NetBSD-8 processes presumably not exiting)
Date: Mon, 20 Apr 2020 21:35:23 +0200

 ----_com.samsung.android.email_5369759761102740
 Content-Type: text/plain; charset=utf-8
 Content-Transfer-Encoding: base64

 Q2FuIGJlIGNsb3NlZC4gSSBleHBlY3QgYmFja3BvcnRpbmcgdG8gYmUgdG9vIG11Y2ggd29yayBh
 bmQgcmlzay5GcmFuawotLS0tLS0tLSBPcmlnaW5hbCBtZXNzYWdlIC0tLS0tLS0tRnJvbTogYWRA
 TmV0QlNELm9yZyBEYXRlOiA0LzIwLzIwICAyMToyMyAgKEdNVCswMTowMCkgVG86IGFkQG5ldGJz
 ZC5vcmcsIG5ldGJzZC1idWdzQG5ldGJzZC5vcmcsIGduYXRzLWFkbWluQG5ldGJzZC5vcmcsIGFk
 QE5ldEJTRC5vcmcsIGthcmRlbEBuZXRic2Qub3JnIFN1YmplY3Q6IFJlOiBrZXJuLzU0MjEwIChO
 ZXRCU0QtOCBwcm9jZXNzZXMgcHJlc3VtYWJseSBub3QgZXhpdGluZykgU3lub3BzaXM6IE5ldEJT
 RC04IHByb2Nlc3NlcyBwcmVzdW1hYmx5IG5vdCBleGl0aW5nU3RhdGUtQ2hhbmdlZC1Gcm9tLVRv
 OiBvcGVuLT5mZWVkYmFja1N0YXRlLUNoYW5nZWQtQnk6IGFkQE5ldEJTRC5vcmdTdGF0ZS1DaGFu
 Z2VkLVdoZW46IE1vbiwgMjAgQXByIDIwMjAgMTk6MjI6NTkgKzAwMDBTdGF0ZS1DaGFuZ2VkLVdo
 eTpzdGFsbGluZyBkdXJpbmcgcG9zdG1hc3RlciBleGl0IHNob3VsZCBiZSBmaXhlZCB3aXRoIHJl
 Y2VudCBwbWFwICYgdXZtIHdvcmsuYmVpbmcgcmVhbGlzdGljIGFib3V0IGl0LCB0aGlzIHdvcmsg
 aXNuJ3QgZ29pbmcgdG8gYmUgYmFja3BvcnRlZCB0byA5LzgvNy5hcmUgeW91IGhhcHB5IHRvIGNs
 b3NlIHRoZSBQUiAoYW5kIHVzZSAtY3VycmVudC8xMCkgb3Igd2FudCBpdCBrZXB0IG9wZW4/

 ----_com.samsung.android.email_5369759761102740
 Content-Type: text/html; charset=utf-8
 Content-Transfer-Encoding: base64

 PGh0bWw+PGhlYWQ+PG1ldGEgaHR0cC1lcXVpdj0iQ29udGVudC1UeXBlIiBjb250ZW50PSJ0ZXh0
 L2h0bWw7IGNoYXJzZXQ9VVRGLTgiPjwvaGVhZD48Ym9keSBkaXI9ImF1dG8iPkNhbiBiZSBjbG9z
 ZWQuIEkgZXhwZWN0IGJhY2twb3J0aW5nIHRvIGJlIHRvbyBtdWNoIHdvcmsgYW5kIHJpc2suPGRp
 diBkaXI9ImF1dG8iPkZyYW5rPC9kaXY+PGRpdj48YnI+PC9kaXY+PGRpdiBzdHlsZT0iZm9udC1z
 aXplOjEwMCU7Y29sb3I6IzAwMDAwMCIgZGlyPSJhdXRvIj48L2Rpdj48ZGl2IHN0eWxlPSJmb250
 LXNpemU6MTAwJTtjb2xvcjojMDAwMDAwIiBkaXI9ImF1dG8iPjwhLS0gb3JpZ2luYWxNZXNzYWdl
 IC0tPjxkaXY+LS0tLS0tLS0gT3JpZ2luYWwgbWVzc2FnZSAtLS0tLS0tLTwvZGl2PjxkaXY+RnJv
 bTogYWRATmV0QlNELm9yZyA8L2Rpdj48ZGl2PkRhdGU6IDQvMjAvMjAgIDIxOjIzICAoR01UKzAx
 OjAwKSA8L2Rpdj48ZGl2PlRvOiBhZEBuZXRic2Qub3JnLCBuZXRic2QtYnVnc0BuZXRic2Qub3Jn
 LCBnbmF0cy1hZG1pbkBuZXRic2Qub3JnLCBhZEBOZXRCU0Qub3JnLCBrYXJkZWxAbmV0YnNkLm9y
 ZyA8L2Rpdj48ZGl2PlN1YmplY3Q6IFJlOiBrZXJuLzU0MjEwIChOZXRCU0QtOCBwcm9jZXNzZXMg
 cHJlc3VtYWJseSBub3QgZXhpdGluZykgPC9kaXY+PGRpdj48YnI+PC9kaXY+PC9kaXY+U3lub3Bz
 aXM6IE5ldEJTRC04IHByb2Nlc3NlcyBwcmVzdW1hYmx5IG5vdCBleGl0aW5nPGJyPjxicj5TdGF0
 ZS1DaGFuZ2VkLUZyb20tVG86IG9wZW4tJmd0O2ZlZWRiYWNrPGJyPlN0YXRlLUNoYW5nZWQtQnk6
 IGFkQE5ldEJTRC5vcmc8YnI+U3RhdGUtQ2hhbmdlZC1XaGVuOiBNb24sIDIwIEFwciAyMDIwIDE5
 OjIyOjU5ICswMDAwPGJyPlN0YXRlLUNoYW5nZWQtV2h5Ojxicj5zdGFsbGluZyBkdXJpbmcgcG9z
 dG1hc3RlciBleGl0IHNob3VsZCBiZSBmaXhlZCB3aXRoIHJlY2VudCBwbWFwICZhbXA7IHV2bSB3
 b3JrLjxicj5iZWluZyByZWFsaXN0aWMgYWJvdXQgaXQsIHRoaXMgd29yayBpc24ndCBnb2luZyB0
 byBiZSBiYWNrcG9ydGVkIHRvIDkvOC83Ljxicj5hcmUgeW91IGhhcHB5IHRvIGNsb3NlIHRoZSBQ
 UiAoYW5kIHVzZSAtY3VycmVudC8xMCkgb3Igd2FudCBpdCBrZXB0IG9wZW4/PGJyPjxicj48YnI+
 PGJyPjwvYm9keT48L2h0bWw+

 ----_com.samsung.android.email_5369759761102740--

State-Changed-From-To: feedback->closed
State-Changed-By: maya@NetBSD.org
State-Changed-When: Mon, 20 Apr 2020 20:13:49 +0000
State-Changed-Why:
frank said:
Can be closed. I expect backporting to be too much work and risk.Frank


>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.46 2020/01/03 16:35:01 leot Exp $
$NetBSD: gnats_config.sh,v 1.9 2014/08/02 14:16:04 spz Exp $
Copyright © 1994-2020 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.