Re: ZFS ARC max sort of not working?
On 07/05/14 07:40, Hans Petter Selasky wrote: On 07/05/14 02:01, Steven Hartland wrote: Related to the global sysctl changes recently perhaps? I'll have a look. The global tunables only support integers and strings, no human shortcuts. This should be easy to fix though. Hi, The following commit should fix the above mentioned problem: http://svnweb.freebsd.org/changeset/base/268285 --HPS ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: keyboard break to debugger broken?
Peter Jeremy wrote this message on Fri, Jul 04, 2014 at 21:27 +1000: On 2014-Jul-04 02:28:48 -0700, John-Mark Gurney j...@funkthat.com wrote: So, I recently tried to break into the debugger w/ the various key sequences that I know about, and none of them worked... I've tried CTRL-ESC, ALT-ESC, CTRL-ALT-ESC, CTRL-PRTSCR, ALT-PRTSCR and CTRL-ALT-PRTSCR, and many other different ones... I've verified that I can sysctl debug.kdb.enter=1 to enter the debugger, and the CTRL-ALT-PAUSE works to suspend the machine, and CTRL-ALT-DEL works to reboot... Does anyone know if this works? It works for me on 10.0. Do you have debug.kdb.break_to_debugger=1 and hw.syscons.kbd_debug=1 (if you're using syscons)? Turns out I didn't... and you didn't need to... emaste helped me the other night discover this.. Apparently rwatson changed this and the docs never got updated... I saw BREAK_TO_DEBUGGER, bug it was documented as being *serial* only, not syscons... I've updated NOTES and added an entry for defaults/loader.conf, but more docs need to be updated that BREAK_TO_DEBUGGER or the tunable or sysctl needs to be set before if works.. -- John-Mark Gurney Voice: +1 415 225 5579 All that I will do, has been done, All that I have, has not. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: keyboard break to debugger broken?
On Jul 5, 2014, at 1:33 PM, John-Mark Gurney j...@funkthat.com wrote: Peter Jeremy wrote this message on Fri, Jul 04, 2014 at 21:27 +1000: On 2014-Jul-04 02:28:48 -0700, John-Mark Gurney j...@funkthat.com wrote: So, I recently tried to break into the debugger w/ the various key sequences that I know about, and none of them worked... I've tried CTRL-ESC, ALT-ESC, CTRL-ALT-ESC, CTRL-PRTSCR, ALT-PRTSCR and CTRL-ALT-PRTSCR, and many other different ones... I've verified that I can sysctl debug.kdb.enter=1 to enter the debugger, and the CTRL-ALT-PAUSE works to suspend the machine, and CTRL-ALT-DEL works to reboot... Does anyone know if this works? It works for me on 10.0. Do you have debug.kdb.break_to_debugger=1 and hw.syscons.kbd_debug=1 (if you're using syscons)? Turns out I didn't... and you didn't need to... emaste helped me the other night discover this.. Apparently rwatson changed this and the docs never got updated... I saw BREAK_TO_DEBUGGER, bug it was documented as being *serial* only, not syscons... I've updated NOTES and added an entry for defaults/loader.conf, but more docs need to be updated that BREAK_TO_DEBUGGER or the tunable or sysctl needs to be set before if works.. I don’t know about 10/11 but 9.3 (and probably late 9.2) introduced console_port which silently overrides the hint.uart.n.flags=0x10 to set which serial port is the ‘console’, in my case I have several platforms (mainly sunfire’s from sun) that only have uart1 available. somehow, i have the feeling that getting the serial console (needed to run serial-over-lab) to work is getting more in the region of magic than technology. please, some of us really need the serial console to work so that we can rescue servers when all else fails. thanks, danny -- John-Mark Gurney Voice: +1 415 225 5579 All that I will do, has been done, All that I have, has not. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 10.0-RELEASE BTX halted on DELL R900
On 05 Jul 2014, at 08:09, Arrigo Marchiori ard...@yahoo.it wrote: On Fri, Jul 04, 2014 at 03:37:27PM +0800, wsk wrote: lists I met a BTX halted problem while upgrade Freebsd 9.0-RC3 to 10.0-Release via freebsd-update. and please check the link below: http://sw.gddsn.org.cn/jopens/test/btx.jpg BTW: I can booted 10.0-R from DVD-ROM as expected but got same error message with flash-driver. I don't remember if that error message means ``division by zero''. It certainly looks a lot like it. The code at cs:eip from the OP's screenshot disassembles to: 36217: f7 35 bc d6 03 00 divl 0x3d6bc 3621d: 85 ff test %edi,%edi 3621f: 74 05 je 0x36226 36221: 89 1f mov%ebx,(%edi) 36223: 89 4f 04mov%ecx,0x4(%edi) 36226: 89 c2 mov%eax,%edx 36228: e9 c2 00 00 00 jmp0x362ef 3622d: 66 c7 45 ea 00 00 movw $0x0,-0x16(%ebp) 36233: 89 c8 mov%ecx,%eax This is a piece of code from /usr/src/lib/libstand/qdivrem.c, which is used to do 64-bit divides. It would be nice if you could try out this loader binary, which has a few additional checks for zero sector counts or sizes: http://www.andric.com/freebsd/loader.edd SHA256 (loader.edd) = 89f99500adb3a8feaa84336ce625975bcfdc0f886514ab02de4992859a671aa9 However, this might still mis-detect your disk sizes, obviously. Just in case, you could try the patch attached to this bug: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=176748 The patch was compiled for 9-STABLE; if it does not apply to the 10.0 sources, then drop me a line so I can adapt it. I tried this patch on a few FreeBSD VMs, and each of them stopped being able to mount the root filesystem because of it. I don't really know what the explanation is... -Dimitry signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [HEADS-UP] Problem with clang in 9-stable [was: r268244 (stable/9) seems to break sysctl hw.ncpu]
On Sat, Jul 05, 2014 at 01:49:44PM +0100, David Chisnall wrote:
On 4 Jul 2014, at 19:18, David Wolfskill da...@catwhisker.org wrote:
clang -O2 -pipe -std=gnu99 -Qunused-arguments -fstack-protector
-Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter
-Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized
-Wno-pointer-sign -Wno-empty-body -Wno-string-plus-int
-Wno-unused-const-variable -Wno-tautological-compare -Wno-unused-value
-Wno-parentheses-equality -Wno-unused-function -Wno-enum-conversion -o
sysctl sysctl.o
This compile line is turning off a lot of warnings. In particular,
-Wno-uninitialized and -Wno-parentheses-equality are likely to hide warnings
that refer to real errors. It sounds like this case was one of them - if
these warnings were on then we'd have got a build failure rather than an
executable that depended on undefined behaviour.
So I checked src/sbin/sysctl/Makefile first; it's fairly vanilla:
# @(#)Makefile8.1 (Berkeley) 6/6/93
# $FreeBSD: stable/9/sbin/sysctl/Makefile 203917 2010-02-15 14:08:06Z
uqs $
PROG= sysctl
WARNS?= 3
MAN=sysctl.8
.include bsd.prog.mk
And the -Wno-uninitialized (at least) comes from bsd.sys.mk:
.if ${WARNS} = 2 ${WARNS} = 4
# XXX Delete -Wuninitialized by default for now -- the compiler doesn't
# XXX always get it right.
CWARNFLAGS+=-Wno-uninitialized
.endif # WARNS =2 WARNS = 4
A bit later, we see the origin of -Wno-parentheses-equality:
# Clang has more warnings enabled by default, and when using -Wall, so if WARNS
# is set to low values, these have to be disabled explicitly.
.if ${COMPILER_TYPE} == clang !defined(EARLY_BUILD)
.if ${WARNS} = 6
CWARNFLAGS+=-Wno-empty-body -Wno-string-plus-int -Wno-unused-const-variable
.endif # WARNS = 6
.if ${WARNS} = 3
CWARNFLAGS+=-Wno-tautological-compare -Wno-unused-value\
-Wno-parentheses-equality -Wno-unused-function
-Wno-enum-conversion
.endif # WARNS = 3
.if ${WARNS} = 2
CWARNFLAGS+=-Wno-switch -Wno-switch-enum -Wno-knr-promoted-parameter
.endif # WARNS = 2
.if ${WARNS} = 1
CWARNFLAGS+=-Wno-parentheses
.endif # WARNS = 1
.if defined(NO_WARRAY_BOUNDS)
CWARNFLAGS+=-Wno-array-bounds
.endif # NO_WARRAY_BOUNDS
.endif # CLANG
I don't know that there's likely to be a huge amount of interest
in addressing the issue for stable/9, but stable/10 looks similar,
and while I see some differences in head, the code in head's
bsd.sys.mk may well be functionally equivalent.
I'm happy to help test if someone wants to put together patches to
(at least) reduce the extent to which we have executables depending
on undefined behavior.
Peace,
david
--
David H. Wolfskill da...@catwhisker.org
Taliban: Evil cowards with guns afraid of truth from a 14-year old girl.
See http://www.catwhisker.org/~david/publickey.gpg for my public key.
pgpclOWC9pp0z.pgp
Description: PGP signature
Re: [HEADS-UP] Problem with clang in 9-stable [was: r268244 (stable/9) seems to break sysctl hw.ncpu]
On 05 Jul 2014, at 14:49, David Chisnall thera...@theravensnest.org wrote: On 4 Jul 2014, at 19:18, David Wolfskill da...@catwhisker.org wrote: clang -O2 -pipe -std=gnu99 -Qunused-arguments -fstack-protector -Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign -Wno-empty-body -Wno-string-plus-int -Wno-unused-const-variable -Wno-tautological-compare -Wno-unused-value -Wno-parentheses-equality -Wno-unused-function -Wno-enum-conversion -o sysctl sysctl.o This compile line is turning off a lot of warnings. In particular, -Wno-uninitialized and -Wno-parentheses-equality are likely to hide warnings that refer to real errors. Interestingly, -Wno-uninitialized has been in bsd.sys.mk since r76861, and the accompanying comment (XXX Delete -Wuninitialized by default for now -- the compiler doesn't always get it right) has never been changed. :-) It is probably time to re-enable that warning after 13 years, at least. I'm not so sure about -Wno-parentheses-equality, because that might give quite a few false positives, especially in old, contributed code. -Dimitry signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [HEADS-UP] Problem with clang in 9-stable [was: r268244 (stable/9) seems to break sysctl hw.ncpu]
On 5 Jul 2014, at 14:07, Dimitry Andric d...@freebsd.org wrote: Interestingly, -Wno-uninitialized has been in bsd.sys.mk since r76861, and the accompanying comment (XXX Delete -Wuninitialized by default for now -- the compiler doesn't always get it right) has never been changed. :-) It is probably time to re-enable that warning after 13 years, at least. It probably only wants enabling for clang. GCC (at least, GCC 4.2.1) performs this analysis based on analyses run by the optimisers and so the warnings are dependent on optimisation level. David ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
[head tinderbox] failure on ia64/ia64
TB --- 2014-07-05 17:09:16 - tinderbox 2.22 running on freebsd-current.sentex.ca TB --- 2014-07-05 17:09:16 - FreeBSD freebsd-current.sentex.ca 9.2-STABLE FreeBSD 9.2-STABLE #0 r263721: Tue Mar 25 09:27:39 EDT 2014 d...@freebsd-current.sentex.ca:/usr/obj/usr/src/sys/GENERIC amd64 TB --- 2014-07-05 17:09:16 - starting HEAD tinderbox run for ia64/ia64 TB --- 2014-07-05 17:09:16 - cleaning the object tree TB --- 2014-07-05 17:09:16 - /usr/local/bin/svn stat --no-ignore /src TB --- 2014-07-05 17:09:44 - At svn revision 268286 TB --- 2014-07-05 17:09:45 - building world TB --- 2014-07-05 17:09:45 - CROSS_BUILD_TESTING=YES TB --- 2014-07-05 17:09:45 - MAKEOBJDIRPREFIX=/obj TB --- 2014-07-05 17:09:45 - MAKESYSPATH=/src/share/mk TB --- 2014-07-05 17:09:45 - PATH=/usr/bin:/usr/sbin:/bin:/sbin TB --- 2014-07-05 17:09:45 - SRCCONF=/dev/null TB --- 2014-07-05 17:09:45 - TARGET=ia64 TB --- 2014-07-05 17:09:45 - TARGET_ARCH=ia64 TB --- 2014-07-05 17:09:45 - TZ=UTC TB --- 2014-07-05 17:09:45 - __MAKE_CONF=/dev/null TB --- 2014-07-05 17:09:45 - cd /src TB --- 2014-07-05 17:09:45 - /usr/bin/make -B buildworld Building an up-to-date bmake(1) [...] cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstFindFrom.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstFirst.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstForEach.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstForEachFrom.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstInit.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstInsert.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -D_PATH_DEFSYSPATH=\.../share/mk:/usr/share/mk\ -I. -I/src/contrib/bmake -DMAKE_NATIVE -std=gnu99 -fstack-protector -Wsystem-headers -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wno-pointer-sign-c /src/contrib/bmake/lst.lib/lstIsAtEnd.c cc -O2 -pipe -DNO_PWD_OVERRIDE -I/src/usr.bin/bmake -DBMAKE_PATH_MAX=1024 -DUSE_META -DMAKE_NATIVE -DHAVE_CONFIG_H -DHAVE_CONFIG_H -DBMAKE_PATH_MAX=1024
