daily CVS update output

Updating src tree: P src/distrib/sets/lists/modules/mi P src/doc/3RDPARTY P src/external/bsd/byacc/bin/yacc.1 P src/external/bsd/libarchive/dist/libarchive/libarchive-formats.5 P src/external/gpl2/groff/dist/tmac/groff_mdoc.man P

Re: gcc 5.3 version of libasan is broken

In article <8e911ed8-0fc3-ab57-6be0-11aa5e23e...@rk.phys.keio.ac.jp>, Rin Okuyama wrote: >Note that -fsanitize=address is not yet working on 32bit big-endian >machines, although it does not fall into SEGV. For example, Report() >does not work as return values of

Re: gcc 5.3 version of libasan is broken

In article <86442cf7-1763-9b78-0c4e-173e4e27e...@rk.phys.keio.ac.jp>, Rin Okuyama wrote: >I found two problems on libasan of gcc 5.3. > >(1) build fails on arm: > >http://releng.netbsd.org/builds/HEAD/201606061330Z/ > >As we use dwarf EH, we must disable __arm__

Re: systems with -current wm(4) hang?

Hi, I'm sorry that I forgot to let you know the fix. On 2016/06/07 3:33, John D. Baker wrote: > Now using if_wm.c r1.411. Thank you for rechecking. I'm glad the revision fixes almost system problem. > The amd64-class systems I've tested so far: > snip > Dell PowerEdge SC430, PCI/PCIe busses:

gmake/dist/main.c

install-info: No such file or directory for /usr/src/obj/tooldir.NetBSD-7.99.30-amd64/info/make.info NOT REBUILDING /usr/src/tools/gmake/../../external/gpl2/gmake/dist/config.h.in NOT REBUILDING /usr/src/tools/gmake/../../external/gpl2/gmake/dist/configure if cc

Re: gcc 5.3 version of libasan is broken

Sorry for bothering you many times. I carelessly considered this is the problem only for 32bit big-endian, but it also affects 64bit machines. A trivial example is the case where a return value of a syscall is int32_t. On 2016/06/07 5:46, Rin Okuyama wrote: Note that -fsanitize=address is not

Re: gcc 5.3 version of libasan is broken

Note that -fsanitize=address is not yet working on 32bit big-endian machines, although it does not fall into SEGV. For example, Report() does not work as return values of __syscall(SYS_write) and strlen(3) are not consistent due to a similar problem to (2). Should we use syscall(2) rather than

gcc 5.3 version of libasan is broken

I found two problems on libasan of gcc 5.3. (1) build fails on arm: http://releng.netbsd.org/builds/HEAD/201606061330Z/ As we use dwarf EH, we must disable __arm__ specific codes in sanitizer_unwind_posix_libcdep.cc, cf. gcc.old version of sanitizer_netbsd.cc:

Re: motg

Additionally, this was committed as src/sys/dev/usb/motg.c 1.17. On Mon, Jun 06, 2016 at 12:38:33PM +0200, Thomas Klausner wrote: > netbsd-bugs is just for the GNATS traffic, please use current-users@ > instead (cc'ed). > > Thanks, > Thomas > > On Sun, Jun 05, 2016 at 01:56:46PM +0300, Artturi

Re: motg

netbsd-bugs is just for the GNATS traffic, please use current-users@ instead (cc'ed). Thanks, Thomas On Sun, Jun 05, 2016 at 01:56:46PM +0300, Artturi Alm wrote: > Hi, > > copypaste bug? > > -Artturi > > --- motg.cSun Jun 5 13:35:13 2016 > +++ motg.cSun Jun 5 13:39:51 2016 > @@