submitter has been refered to the ISC to determine if there his problems
are due to a bug in the script.
As a side note, `man 8 ifconfig' does not show "ifconfig -alias" as being
acceptable, but it is and does server a purpose.
to generate the special .type and .size directives which are used to set
the corresponding fields of the linker symbol table entries in the ELF
object file. As such they are not i386-specific and thus belong in our
MI header. Otherwise on the Alpha we don't properly give the type and
size of dynamic symbols. Bintuil versions past 2.9.1 warn of this and
w/o this change, `ld' generates a lot of warnings during a `make world'.
it about the FreeBSD equivalent) This version of that patch selects the
Linux code only if the case is Linux. This puts the Linux code on par with
the FreeBSD code.
for function. For example: for stat(2) syscall (owned by libc) we will add
following code:
.Sh LIBRARY
.Lb libc
and it will be rendered to:
LIBRARY
Standard C Library (libc, -lc)
Suggested by: mpp
Idea obtained from: NetBSD
With jb's rev 1.2 commit to usr.bin/lorder/lorder.sh we don't depend on
the 4.4BSD's ``nm -o'' behavior.
Previous to 4.3BSD-Reno, ``nm -o'' did not output the .o's filename on a
line by itself. The change occurred between CSRG's nm.c rev 4.8 (1987) and
rev 5.1 (in 1989), which was "new version from Hans Huebner hans@garp.mit.edu,
huebner@db0tui6.BITNET". The Binutils maintainers would rather cater to
a loud boisterous user of 4.3BSD VAXen which has its own native toolchain,
rather than a modern Unix with multiple orders of magnitude more users
and in which Binutils *is* the native toolchain.
I did not catch this on the EGCS 1.1.x --> GCC 2.95 upgrade.
So propogate this change to our custom hacks.
PR: 15549
Submitted by: Naohiko Tsuji <yakisoba@osaka.interq.or.jp>
Patrick Bihan-Faou <patrick@mindstep.com>
maintainers.
After we established our branding method of writing upto 8 characters of
the OS name into the ELF header in the padding; the Binutils maintainers
and/or SCO (as USL) decided that instead the ELF header should grow two new
fields -- EI_OSABI and EI_ABIVERSION. Each of these are an 8-bit unsigned
integer. SCO has assigned official values for the EI_OSABI field. In
addition to this, the Binutils maintainers and NetBSD decided that a better
ELF branding method was to include ABI information in a ".note" ELF
section.
With this set of changes, we will now create ELF binaries branded using
both "official" methods. Due to the complexity of adding a section to a
binary, binaries branded with ``brandelf'' will only brand using the
EI_OSABI method. Also due to the complexity of pulling a section out of an
ELF file vs. poking around in the ELF header, our image activator only
looks at the EI_OSABI header field.
Note that a new kernel can still properly load old binaries except for
Linux static binaries branded in our old method.
*
* For a short period of time, ``ld'' will also brand ELF binaries
* using our old method. This is so people can still use kernel.old
* with a new world. This support will be removed before 5.0-RELEASE,
* and may not last anywhere upto the actual release. My expiration
* time for this is about 6mo.
*
-fpic code that damages symbol locations at runtime.
The only know occurance in our tree (src and ports) was locatime.c,
which was just changed to generate code that doesn't trigger the
problem.
This is a workaround, the real cause is that our gas doesn't
understand code our gcc generates for some -O -fpic code. They are
expected to be back in sync soon, but until then (including
4.0-RELEASE) we need to prevent people from using bad -fpic code.
PR: avoids such things as in bin/16862
Submitted by: bde
Approved by: jkh
I had resisted this for a while because I didn't want to take these
file off the vendor branch. But it turns out they were already off
the vendor branch anyway.
PR: gnu/9987
Submitted by: Vladimir Kushnir <kushn@mail.kar.net>
Approved by: jkh
Sorry there were still several bugs.
-error retry at af missmatch was incomplete.
-af matching for source addr option was wrong
-socket was not freed at retry.
Approved by: jkh
-Should retry as much as possible when some of source
routing intermediate hosts' address families missmatch
happened.
(such as when a host has only A record, and another host
has each of A and AAAA record.)
-Should retry as much as possible when dest addr and
source addr(specified with -s option) address family
missmatch happend
Approved by: jkh
back to the vendor branch.
Re-merge changes from the vendor branch.
Undo the local spammed changes that I could find. There are probably
more local fixes that were clobbered that I've missed.
Now when tcp_wrapper is enabled by inetd -wW,
several accesses which should be permitted are refused only for IPv6,
if hostname is used to decide the host to be allowed.
IPv6 users will be just upset.
About security related concern.
-All extensions are wrapped by #ifdef INET6, so people can completely
disable the extension by recompile libwrap without INET6 option.
-Access via IPv6 is not enabled by default.
People need to enable IPv6 access by changing /etc/inetd.conf at first,
by adding tcp6 and/or tcp46 entries.
-The base of patches are from KAME package and are actually daily used
for more than a year in several Japanese IPv6 environments.
-Patches are reviewed by markm.
Approved by: jkh
Submitted by: Hajimu UMEMOTO <ume@mahoroba.org>
Reviewed by: markm
Obtained from: KAME project
repository (dated 5-3-1999).
This fixes the endless loop the assembler gets in when it is trying to
decide how big part of the exception handling table should be on the Alpha.
This version of this file allows qradiobutton.cpp from qt and widget.cc
from kdelibs11 to be compilable.
getnameinfo() don't return error at name resolving failure.
But it is used at doaddrlookup(-N) case in telnet, error need to be
returned to correctly initialize hostname buffer.
Discovered at checking recent KAME repository change, noticed by itojun.
ICANON is off is implementation-defined. Under BSD, IEXTEN is
independent of ICANON, so it must be turned off to get "raw" mode.
This was first fixed in rev.1.4 (1995/10/21) of libncurses/lib_raw.c,
but the change was lost in the downgrade to the contrib version.
The fix here is the same as in the old rev.1.4, less style bugs. A
better fix would use cfmakeraw(3) to actually handle all of the
complications for switching to raw mode.
"The problem is that egcs/gcc-2.95's reorganisation of the prologue and
epilogue code to use rtl instead of output_asm_insn() completely broke our
hooks. rtl is emitted in a different order, only after optimisation, while
output_asm_insn() is emitted immediately. rtl is presumably used so that
the prologue and epilogue can be optimised.
I couldn't find any good examples to copy. gcc's own
FUNCTION_BLOCK_PROFILER still uses output_asm_insn() and seems to be
completely broken. One of the XXX comments points to this.
IIRC, the hacks here basically arrange to emit magic label names; then when
the magic names are output, they are transformed into prologue and epilogue
code."
Submitted by: bde
page of -mdoc manual pages in troff mode. The problem has been
introduced in rev 1.12; apparently ending a line inside a macro
definition with \c doesn't behave as Bill expected, and troff was
fooled to believe previous output was already pending thus a new page
needed to be emitted. Using a troff comment .\" instead of \c seems
to solve this, alas the brain^Wnice sed(1) magic needs to be taught to
_not_ strip those comments now.
I think -stable is not affected by this.
