o Up to 8 arguments are allowed. This is the number of arguments
passed in registers. Subsequent registers are passed on the stack.
Trying to deal with this is not easy in C and likely forces us to
use assembly code. Let's avoid that for now. There's no indication
that more than 8 arguments is a strong requirement (Linux also has
an 8 argument limit).
o We expect that the stack base is 16-byte aligned and the stack
size is a multiple of 16-byte. We bomb out if this is not the case.
We probably want to be less strict by enforcing it ourselves. For
now it's better to not hide gross alignment bogons by silently
correcting it.
to a buffer in the big key/data case, memmove() was used on pointers
to size_ts, but only sizeof(u_int32_t) bytes where copied. This broke
on big_endian architectures where sizeof(size_t) > sizeof(u_int32_t).
This bug broke portupgrade (by way of ruby_bdb1) on sparc64.
Approved by: re (rwatson)
are not initialized at this place. Move the initializing
before the non-blocking check.
Submitted by: Marius Strobl <marius@alchemy.franken.de>
Reviewed by: jhb
Approved by: re
just read() in non-blocking mode too. The reason is obvious. NetBSD
uses a complete different way to get the credentials so this patch
only applies to FreeBSD.
Reviewed by: rwatson
Approved by: re
Remove the special treatment of non-blocking mode in
the "look ahead function" xdrrec_eof(). It currently
assumes that the last read() in a row of several reads
does not have zero lenght. If this is the case, svc_vc_stat()
does return XPRT_MOREREQS, and the RPC-request aborts because
there is no data to read anymore.
To fix this, go back to the original version of the code
for non-blocking mode until NetBSD comes up with another
possible fix like this one in xdrrec_eof()
if (rstrm->last_frag && rstrm->in_finger == rstrm->in_boundry) {
return TRUE;
}
Return always FALSE in set_input_fragment() for non-blocking
mode. Since this was not used in FreeBSD, I omitted it at the
first time. Now we use this function and we should always
return FALSE for it.
Reviewed by: rwatson
Approved by: re
prime objectives are:
o Implement a syscall path based on the epc inststruction (see
sys/ia64/ia64/syscall.s).
o Revisit the places were we need to save and restore registers
and define those contexts in terms of the register sets (see
sys/ia64/include/_regset.h).
Secundairy objectives:
o Remove the requirement to use contigmalloc for kernel stacks.
o Better handling of the high FP registers for SMP systems.
o Switch to the new cpu_switch() and cpu_throw() semantics.
o Add a good unwinder to reconstruct contexts for the rare
cases we need to (see sys/contrib/ia64/libuwx)
Many files are affected by this change. Functionally it boils
down to:
o The EPC syscall doesn't preserve registers it does not need
to preserve and places the arguments differently on the stack.
This affects libc and truss.
o The address of the kernel page directory (kptdir) had to
be unstaticized for use by the nested TLB fault handler.
The name has been changed to ia64_kptdir to avoid conflicts.
The renaming affects libkvm.
o The trapframe only contains the special registers and the
scratch registers. For syscalls using the EPC syscall path
no scratch registers are saved. This affects all places where
the trapframe is accessed. Most notably the unaligned access
handler, the signal delivery code and the debugger.
o Context switching only partly saves the special registers
and the preserved registers. This affects cpu_switch() and
triggered the move to the new semantics, which additionally
affects cpu_throw().
o The high FP registers are either in the PCB or on some
CPU. context switching for them is done lazily. This affects
trap().
o The mcontext has room for all registers, but not all of them
have to be defined in all cases. This mostly affects signal
delivery code now. The *context syscalls are as of yet still
unimplemented.
Many details went into the removal of the requirement to use
contigmalloc for kernel stacks. The details are mostly CPU
specific and limited to exception_save() and exception_restore().
The few places where we create, destroy or switch stacks were
mostly simplified by not having to construct physical addresses
and additionally saving the virtual addresses for later use.
Besides more efficient context saving and restoring, which of
course yields a noticable speedup, this also fixes the dreaded
SMP bootup problem as a side-effect. The details of which are
still not fully understood.
This change includes all the necessary backward compatibility
code to have it handle older userland binaries that use the
break instruction for syscalls. Support for break-based syscalls
has been pessimized in favor of a clean implementation. Due to
the overall better performance of the kernel, this will still
be notived as an improvement if it's noticed at all.
Approved by: re@ (jhb)
Fixes a problem where directory entries could show up twice: once
on the top layer of the union stack, and once on the bottom layer.
Approved by: re (rwatson)
Those who really need this information can find it in the include file.
* Include a succinct description of the st_birthtime field.
Approved by: re (bmah)
- strip out the nasty PIC_PROLOGUE/EPILOGUE stuff, since we dont have
to lose a register in PIC mode anymore (we use %rip-relative addressing).
- update for C register argument passing conventions.
- convert 32 bit to 64 bit register sizes etc.
Note that the syscall instruction clobbers %rcx, which is inconvenient
because it is the fourth syscall argument, so we use %r10 (another scratch
register) for the 4th syscall arg instead (I picked %r10 to be the same as
NetBSD). int 0x80 is still possible though, and it uses %rcx as usual.
Note that the syscall style syscall does *NOT* preserve all the registers,
unlike int 0x80. We do not preserve the scratch registers except for
%rdi and %rsi. int 0x80 does preserve everything but the return values.
technique) so that we don't wind up calling into an application's
version if the application defines them.
Inspired by: qpopper's interfering and buggy version of strlcpy
setgrent, and endgrent also. (The previous NSS implementation used to
simply twiddle the internal data of the various modules directly.)
A symptom (group list set incorrectly in sshd) was
Reported by: Glenn Johnson <gjohnson@srrc.ars.usda.gov>
Sponsored by: DARPA, Network Associates Laboratories
idea to re-initialize `struct passwd', because e.g. pw_class might
get set by one module, but not by another. Add another call to the
internal pwd_init function to accomplish this.
Sponsored by: DARPA, Network Associates Laboratories
Correct a bug that should have wreaked havoc everywhere, but for
some reason only bit unlucky people who use `-march' optimizations.
The compiler cannot assist one in distinguishing between the two
function calls below.
int nsdispatch(void *, ...);
void *discard;
nsdispatch(&discard, ...); /* correct .. no, really! */
nsdispatch(discard, ...); /* Boom */
Robin provided me with a debugging environment in which I could see
what was going on.
Badness when using CPUTYPE was
Reported by: "Robin P. Blanchard" <Robin.Blanchard@gactr.uga.edu>
Reported by: nork
Sponsored by: DARPA, Network Associates Laboratories
An incorrectly-sized allocation was being made due to an incorrect
argument to the `sizeof' operator. Obvious, because it violated the
`foo = malloc(sizeof(*foo))' idiom. Hard-to-see, because it was a
missing `*' (`*p' versus `**p').
Resulting failure was
Reported by: ache
Sponsored by: DARPA, Network Associates Laboratories
privileges. To do so may cause the NIS server to log spurious and
annoying `access denied' messages.
Reported by: Philip Paeps <philip@paeps.cx>
Sponsored by: DARPA, Network Associates Laboratories
in the mac.3 library man page. They were already cross-referenced
at the end of the man page, just not explicitly listed here.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
would result in an incorrectly terminated grouplist.
login(1) crashes
Reported by: Morten Rodal <morten@rodal.no>,
Matthias Schuendehuette <msch@snafu.de>
string was an incredibly dumb idea (of course it will be changed by an
NSS module on success!). Use a static empty string instead.
Sponsored by: DARPA, Network Associates Laboratories
may not fill in all fields, and in the case of string fields, this could
cause trouble for applications. (The only likely example is `pw_class',
because this field is not used by all modules in all cases.)
Move initialization of struct passwd from module-specific code to the
dispatch code.
The problem of a NULL pw_class was
Noticed by: Philip Paeps <philip@paeps.cx>
and the c^Htrusty ssh(1) command.
Déjà vu by: getpwent.c revision 1.56
Sponsored by: DARPA, Network Associates Laboratories
at least one consumer outside of libc and pwd_mkdb.
Adjust the versioning in libc and pwd_mkdb accordingly.
named was the application affected, and that fact was first
Reported by: Zherdev Anatoly <tolyar@mx.ru>
Sponsored by: DARPA, Network Associates Laboratories
family of functions using the new nsdispatch(3) core. Remove
arbitrary size limits when using the thread-safe versions.
= Re-implement the traditional getpwent(3)/getgrent(3) functions on
top of the thread-safe versions.
= Update the on-disk format of the hashed version of the passwd(5)
databases to allow for versioned entries. The legacy version is
`3'. (Don't ask.)
= Add support for version `4' entries in the passwd(5) database.
Entries in this format are identical to version 3 entries except
that all integers are stored as 32-bit integers in network byte
order (big endian).
= pwd_mkdb is updated to generate both version 3 and version 4
entries.
Sponsored by: DARPA, Network Associates Laboratories
may be built into libc (`static NSS modules') or dynamically loaded
via dlopen (`dynamic NSS modules'). Modules are loaded/initialized
at configuration time (i.e. when nsdispatch is called and nsswitch.conf
is read or re-read).
= Make the nsdispatch(3) core thread-safe.
= New status code for nsdispatch(3) `NS_RETURN', currently used to
signal ERANGE-type issues.
= syslog(3) problems, don't warn/err/abort.
= Try harder to avoid namespace pollution.
= Implement some shims to assist in porting NSS modules written for
the GNU C Library nsswitch interface.
Sponsored by: DARPA, Network Associates Laboratories
they resemble one another, but POSIX.1e interfaces were not sufficiently
expressive to do what we needed.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
%f and sufficiently short %g specifiers where the precision was
explicitly zero, no '#' flag was specified, and the floating point
argument was > 0 and <= 0.5. While at it, add some comments to better
explain the relevant bits of code.
Noticed by: Christoph Kukulies <kuku@physik.rwth-aachen.de>
by allprison_mtx), a unique prison/jail identifier field, two path
fields (pr_path for reporting and pr_root vnode instance) to store
the chroot() point of each jail.
o Add jail_attach(2) to allow a process to bind to an existing jail.
o Add change_root() to perform the chroot operation on a specified
vnode.
o Generalize change_dir() to accept a vnode, and move namei() calls
to callers of change_dir().
o Add a new sysctl (security.jail.list) which is a group of
struct xprison instances that represent a snapshot of active jails.
Reviewed by: rwatson, tjr
be printed.
- Fix %f conversions where the number of significant digits is < expt.
This would be a one-line change were it not for thousands separators.
Noticed by tjr.
- Remove some unnecessary code in the parsing of precision specifiers.
- We used to round long double arguments to double. Now we print
them properly.
- Bugs involving '%F', corner cases of '#' and 'g' format
specifiers, and the '.*' precision specifier have been
fixed.
- Added support for the "'" specifier to print thousands' grouping
characters in a locale-dependent manner.
- Implement the __vfprintf() side of hexadecimal floating point
support. All that is still needed is a routine to convert the
mantissa to hex digits one nibble at a time in the style of ultoa().
Reviewed by: silence on standards@
- __vfprintf()'s 'buf' has never been used for floating point, so
don't define it in terms of (incorrect) constants describing
floating point numbers. The actual size needed depends on
sizeof(uintmax_t) and locale details, so I slightly overestimated.
- We don't need a 308-character buffer to store the string "308".
With long doubles and %a we need more than three characters, though.
strange things might happen when garbage values in the struct
get passed in to localtime_r() and family.
Noticed by: marcus
Approved by: markm (mentor)(implicit)
critical and should not be killed when pageout is looking for more
memory pages in all the wrong places.
Reviewed by: arch@
Sponsored by: St. Bernard Software
from strptime(3). Previously, they would get filled only
for the %s specifier and as a side effect of using the
the %Z specifier with a GMT time zone.
PR: misc/48993
Approved by: markm (mentor)
Silence on: -standards
more complicated things than just setting the lock to 0.
- Implement stubs for this function in libc and the two threading libraries
that are currently in the tree.
by NIS work, like nsswitch.conf(5) promises to be able to.
(These modifications will be fed back to NetBSD, of course)
- In endusershell(), do not set `sl' to NULL if we know it already has
that value.
dtoa() is buggy. The bug would cause incorrect output to be
generated when format strings such as '%5.0f' were used with
nonzero numbers whose magnitude is less than 1.
Reported by: df(1) by way of periodic(8)
Reviewed by: mike
package, a more recent, generalized set of routines. Among the
changes:
- Declare strtof() and strtold() in stdlib.h.
- Add glue to libc to support these routines for all kinds
of ``long double''.
- Update printf() to reflect the fact that dtoa works slightly
differently now.
As soon as I see that nothing has blown up, I will kill
src/lib/libc/stdlib/strtod.c. Soon printf() will be able
to use the new routines to output long doubles without loss
of precision, but numerous bugs in the existing code must
be addressed first.
Reviewed by: bde (briefly), mike (mentor), obrien
not save (restore) the global pointer (GP) in the jmpbuf in setjmp
(longjmp) because it's not needed in general. GP is considered a
scratch register at callsites and hence is always restored after a
call (when it's possible that the call resolves to a symbol in a
different loadmodule; otherwise GP does not have to be saved and
restored at all), including calls to setjmp/longjmp. There's just
one problem with this now that we use setjmp/longjmp for context
switching: A new context must have GP defined properly for the
thread's entry point. This means that we need to put GP in the
jmpbuf and consequently that we have to restore is in longjmp.
This automaticly requires us to save it as well.
When setjmp/longjmp isn't used for context switching, this can be
reverted again.
the J_SIG0 field. While here, rename J_SIG0 to J_SIGSET and
remove J_SIG1. The main reason for this change is that the
128-bit sigset_t is now aligned on a 16-byte boundary, which
allows us to use 16-byte atomic loads and stores on CPUs that
support it. The removal of J_SIG1 is done to avoid confusion:
it is never accessed and should not be. Renaming J_SIG0 to
J_SIGSET is the icing on the cake that's better done now than
later.
The background info in this man page needs rewriting
in some parts since the last major changes
to the code, however it still accuratly reflects how to use the
API.
* use correct error detection of realloc failure
* strtol negative return check
* use strtol to validate string instead of rolling our own
validation code
* terminate the command sequence correctly
seed->first value correlation. It breaks rand_r()... Other possible methods
like shuffling inside aray will breaks rand_r() too, because it assumes
only one word state, i.e. nothing extra can be added after seed assignment
in srand().
BTW, for old formulae seed->first value correlation is not so monotonically
increased as with other Linear Congruential Generators of this type only
becase arithmetic overflow happens. But overflow affects distribution
and lower bits very badly, as many articles says, such type of overflow
not improves PRNG.
So, monotonically increased seed->first value correlation problem remains...
Only warnings that could be fixed without changing the generated object
code and without restructuring the source code have been handled.
Reviewed by: /sbin/md5
Introdice RTLD_SELF special handle and properly process it within
dlsym() and dlinfo() functions.
The intention is to improve our compatibility with Solaris and
to make a Java port easier.
Partially submitted by: phantom
isnormal(). The current isinf() and isnan() are perserved for
binary compatibility with 5.0, but new programs will use the macros.
o Implement C99 comparison macros isgreater(), isgreaterequal(),
isless(), islessequal(), islessgreater(), isunordered().
Submitted by: David Schultz <dschultz@uclink.Berkeley.EDU>
o Add a MD header private to libc called _fpmath.h; this header
contains bitfield layouts of MD floating-point types.
o Add a MI header private to libc called fpmath.h; this header
contains bitfield layouts of MI floating-point types.
o Add private libc variables to lib/libc/$arch/gen/infinity.c for
storing NaN values.
o Add __double_t and __float_t to <machine/_types.h>, and provide
double_t and float_t typedefs in <math.h>.
o Add some C99 manifest constants (FP_ILOGB0, FP_ILOGBNAN, HUGE_VALF,
HUGE_VALL, INFINITY, NAN, and return values for fpclassify()) to
<math.h> and others (FLT_EVAL_METHOD, DECIMAL_DIG) to <float.h> via
<machine/float.h>.
o Add C99 macro fpclassify() which calls __fpclassify{d,f,l}() based
on the size of its argument. __fpclassifyl() is never called on
alpha because (sizeof(long double) == sizeof(double)), which is good
since __fpclassifyl() can't deal with such a small `long double'.
This was developed by David Schultz and myself with input from bde and
fenner.
PR: 23103
Submitted by: David Schultz <dschultz@uclink.Berkeley.EDU>
(significant portions)
Reviewed by: bde, fenner (earlier versions)
to remove part of seed -> 1st value correlation. Correlation still remains
because of algorithm limits. Note that old algorithm have even stronger
correlation, especially in the lower bits area, but not eye-visible, as
current one.
at 0 as designed. Its BSD adaptation tries to fight it by mapping 0 to
2147483647 after calculation, but this method not works since 2147483647
seed returns to 0 again on the next interation. Instead of after calculation
mapping, map 0 to another value _before_ calculation, so it never stucks.
checks, including the "open directory" check or a MAC check fail,
after the working directory of the process has been changed, then
the cwd of the process will be left as the target directory rather
than the original directory.
At some point, this bug might be fixable by performing the directory
change only after permission is granted for the change. In the
mean time document it (it's been there for a while).
is finished. This fixes clients doing two RPCs over the
same connection at the same time. Without this fix, we
could end with a reply to old data.
Submitted by: Frank van der Linden <fvdl@netbsd.org>
Reviewed by: rwatson
Obtained from: NetBSD
a NULL filename argument allows a stream's mode to be changed. At the
moment it just recycles the old file descriptor instead of storing the
filename somewhere and using that to reopen the file, as the standard
seems to require. Strictly conforming C99 applications probably can't
tell the difference but POSIX ones can.
PR: 46791
to be called on first sight of trouble.
"sensitive" is somewhat arbitrarily defined as "setuid, setgid, uid == root
or gid == wheel".
The 'A' option carries no performance penalty.
It is not possible to override this setting: fix the program instead.
Absentmindedly nodded OK to by: various
o Link mac_get_pid.3 to mac_get.3
o Update SEE ALSO to refer to mac_prepare, and added missing references
o Remove clause #3 on my work
o Update mac_get.3 for the updated MAC API
Sponsored by: DARPA, Network Associates Laboratories
Obtained from: TrustedBSD Project
so that multiple opens of the same semaphore without an intervening
sem_close() return the same object, and so that sem_close() does not
segfault while trying to remove the item from the list.
putchar_unlocked(), putc_unlocked(), feof_unlocked(), ferror_unlocked(),
clearerr_unlocked(), and fileno_unlocked(). The first four are required
by POSIX. The rest are provided for consistency.
