Re: CVS commit: src
On Tue, May 03, 2011 at 06:28:46PM +, David Young wrote: Module Name: src Committed By:dyoung Date:Tue May 3 18:28:46 UTC 2011 Modified Files: src/distrib/sets/lists/comp: mi src/sys/dist/pf/net: pf.c src/sys/netinet: Makefile files.netinet in_pcb.c in_pcb.h in_pcb_hdr.h tcp_input.c tcp_subr.c tcp_usrreq.c tcp_var.h udp_usrreq.c src/sys/netinet6: in6_pcb.c in6_pcb.h in6_src.c ip6_input.c raw_ip6.c udp6_usrreq.c src/sys/rump/net/lib/libnetinet: Makefile.inc src/usr.bin/netstat: Makefile inet.c inet6.c main.c netstat.h Added Files: src/sys/netinet: tcp_vtw.c tcp_vtw.h src/usr.bin/netstat: vtw.c vtw.h Log Message: Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). [...] This changeset broke my qemu test environment. With DIAGNOSTIC, DEBUG, and LOCKDEBUG boot gets as far as starting init; then it sits there for a while and after a few moments prints UVM: pid 2 (sh), uid 0 killed: out of swap Out of memory allocating ksiginfo for pid 2 several times (that is, it kills pid 2 several times, which is itself odd) and then asserts at line 2383 of pmap.c, which is in pmap_destroy checking pmap.pm_obj[i].uo_npages == 0. GENERIC is less overtly broken, but hangs during multiuser boot sometime after printing Adding interface aliases:. This is on amd64 and it appears to be qemu-specific, or possibly specific to my qemu setup; at least nobody else seems to have hit it. It's also possible that it only happens with a new kernel and older userland; the userland in the VM image is 5.99.45 from mid-February. Suggestions? I can try to apply the changes piecemeal to see at what point things blow up, but I'm not familiar enough with the TCP code to tackle it properly without putting a big chunk of time into it, which I don't really have... -- David A. Holland dholl...@netbsd.org
Re: CVS commit: src/tests/lib/libposix
On 5/16/11 1:03 AM, Christos Zoulas wrote: Module Name:src Committed By: christos Date: Mon May 16 00:03:36 UTC 2011 Modified Files: src/tests/lib/libposix: t_rename.c Log Message: h_macros need strlcat and random ugh, please someone remove this header. define _NETBSD_SOURCE so those are defined. The original intent of h_macros was as a place to provide NetBSD-specific check macros that can't be part of the generic codebase of atf. I'm not sure what it currently includes, but I presume it does not follow this rule any more. I'll check later (FSVO later ;-) what pieces could be now moved to atf to clean this up a bit.
Re: CVS commit: src/sys
On Mon, May 16, 2011 at 06:35:35PM +0200, Hauke Fath wrote: Module Name:src Committed By: rmind Date: Sat May 14 17:57:05 UTC 2011 Modified Files: src/sys/kern: kern_resource.c src/sys/sys: resource.h resourcevar.h Log Message: - Sprinkle __read_mostly, consitify maxdmap and maxsmap. This breaks the sparc build, which sets 'maxdmap', see http://nxr.netbsd.org/search?q=project=srcdefs=refs=maxdmappath=hist=. Reminds me of http://www.netbsd.org/ports/ -- remember sparc is tier 2 aka Go fix it yourself if you don't hurt me...
Re: CVS commit: src/sys
On Mon, May 16, 2011 at 08:18:08PM +0200, Martin Husemann wrote: On Mon, May 16, 2011 at 07:29:26PM +0200, Martin S. Weber wrote: Reminds me of http://www.netbsd.org/ports/ -- remember sparc is tier 2 aka Go fix it yourself if you don't hurt me... Well, but the fix is simple in this case: turn it into a MD macro like MAXDMAP() and allow other archs to provide a const implementation, while sparc pays the price and has to fetch it from a variable. or just hack it brutally ( cvs rdiff -u -r1.308 -r1.309 src/sys/arch/sparc/sparc/machdep.c ) -- who cares about hacks in tier 2 architectures when the netbsd goals (http://www.netbsd.org/about/system.html - well designed) only apply to tier 1 ports? (the goals page should be updated anyways to exclude examples such as EISA, TurboChannel, the alpha or pmap port etc) /me shrugs
Re: CVS commit: src/sys
Martin S. Weber ephae...@gmx.net wrote: On Mon, May 16, 2011 at 08:18:08PM +0200, Martin Husemann wrote: On Mon, May 16, 2011 at 07:29:26PM +0200, Martin S. Weber wrote: Reminds me of http://www.netbsd.org/ports/ -- remember sparc is tier 2 aka Go fix it yourself if you don't hurt me... Well, but the fix is simple in this case: turn it into a MD macro like MAXDMAP() and allow other archs to provide a const implementation, while sparc pays the price and has to fetch it from a variable. or just hack it brutally ( cvs rdiff -u -r1.308 -r1.309 src/sys/arch/sparc/sparc/machdep.c ) -- who cares about hacks in tier 2 architectures when the netbsd goals (http://www.netbsd.org/about/system.html - well designed) only apply to tier 1 ports? (the goals page should be updated anyways to exclude examples such as EISA, TurboChannel, the alpha or pmap port etc) The *modification* of those variables in such way is rather a hack. Note that I have asked sparc guys whether they are happy with such change. Tell me more about well designed. -- Mindaugas
Re: CVS commit: src/lib/libc/stdlib
Christos Zoulas chris...@netbsd.org writes: Module Name: src Committed By: christos Date: Fri May 13 23:11:00 UTC 2011 Modified Files: src/lib/libc/stdlib: jemalloc.c malloc.c Log Message: don't let readlink trash errno.; To generate a diff of this commit: cvs rdiff -u -r1.22 -r1.23 src/lib/libc/stdlib/jemalloc.c cvs rdiff -u -r1.52 -r1.53 src/lib/libc/stdlib/malloc.c - Is the existing guard in malloc_init() of malloc.c insufficient? - In the malloc_init_hard() of jemalloc.c, there are another calls may set errno. Especially, the code explicitly handles an error from sysctl(3). enami.