Corinna Vinschen [Mon, 23 Mar 2020 11:13:35 +0000 (12:13 +0100)]
Cygwin: serial: tcsetattr: set timeouts unconditionally
tcsetattr checks if the VTIME and VMIN values changed and only
calls SetCommTimeouts if so. That's a problem if tcsetattr
is supposed to set VTIME and VIMN to 0, because these are the
start values anyway. But this requires to set ReadIntervalTimeout
to MAXDWORD, which just doesn't happen.
Fix this by dropping the over-optimization of checking the old
values before calling SetCommTimeouts,
Corinna Vinschen [Wed, 18 Mar 2020 20:14:06 +0000 (21:14 +0100)]
Cygwin: serial: revamp overlapped IO in read and select
Get rid of WaitCommEvent and using overlapped_armed to share the
same overlapped operation between read and select. Rather, make
sure to cancel the overlapped IO before leaving any of these functions.
Corinna Vinschen [Tue, 17 Mar 2020 16:45:05 +0000 (17:45 +0100)]
Cygwin: serial: read: revamp raw_read, change vmin_ and vtime_ to cc_t
- Datatypes were incorrect, especially vmin_ and vtime_.
Change them to cc_t, as in user space.
- Error checking had a gap or two. Debug output used the
wrong formatting.
- Don't use ev member for ClearCommError and WaitCommEvent.
Both returned values are different (error value vs. event
code). The values are not used elsewhere so it doesn't make
sense to store them in the object. Therefore, drop ev member.
- Some variable names were not very helpful. Especially using
n as lpNumberOfBytesTransferred from GetOverlappedResult and
then actually printing it as if it makes sense was quite
puzzeling.
- Rework the loop and the definition of minchars so that it
still makes sense when looping.
Corinna Vinschen [Tue, 17 Mar 2020 16:24:56 +0000 (17:24 +0100)]
Cygwin: serial: select: simplify peek_serial
- Don't use ev member for ClearCommError and WaitCommEvent.
Both returned values are different (error value vs. event
code). The values are not used elsewhere so it doesn't make
sense to store them in the object.
- Drop local variable ready which is used inconsequentially.
- Since WFSO already waits 10 ms, don't wait again if no char
is in the inbound queue.
- Avoid else if chains.
- Only print one line of debug output on error.
- Drop overlapped_armed < 0 check. This value is only set in
fhandler_serial::raw_read if VTIME > 0, and even then it's only
set to be immediately reset to 0 before calling ReadFile. So
overlapped_armed is never actually < 0 when calling select.
Fabian Schriever [Tue, 17 Mar 2020 14:48:44 +0000 (15:48 +0100)]
Fix for k_tan.c specific inputs
This fix for k_tan.c is a copy from fdlibm version 5.3 (see also
http://www.netlib.org/fdlibm/readme), adjusted to use the macros
available in newlib (SET_LOW_WORD).
This fix reduces the ULP error of the value shown in the fdlibm readme
(tan(1.7765241907548024E+269)) to 0.45 (thereby reducing the error by
1).
This issue only happens for large numbers that get reduced by the range
reduction to a value smaller in magnitude than 2^-28, that is also
reduced an uneven number of times. This seems rather unlikely given that
one ULP is (much) larger than 2^-28 for the values that may cause an
issue. Although given the sheer number of values a double can
represent, it is still possible that there are more affected values,
finding them however will be quite hard, if not impossible.
We also took a look at how another library (libm in FreeBSD) handles the
issue: In FreeBSD the complete if branch which checks for values smaller
than 2^-28 (or rather 2^-27, another change done by FreeBSD) is moved
out of the kernel function and into the external function. This means
that the value that gets checked for this condition is the unreduced
value. Therefore the input value which caused a problem in the
fdlibm/newlib kernel tan will run through the full polynomial, including
the careful calculation of -1/(x+r). So the difference is really whether
r or y is used. r = y + p with p being the result of the polynomial with
1/3*x^3 being the largest (and magnitude defining) value. With x being
<2^-27 we therefore know that p is smaller than y (y has to be at least
the size of the value of x last mantissa bit divided by 2, which is at
least x*2^-51 for doubles) by enough to warrant saying that r ~ y. So
we can conclude that the general implementation of this special case is
the same, FreeBSD simply has a different philosophy on when to handle
especially small numbers.
Corinna Vinschen [Mon, 16 Mar 2020 09:20:16 +0000 (10:20 +0100)]
Cygwin: serial: wait for CancelIo completion
Per https://devblogs.microsoft.com/oldnewthing/20110202-00/?p=11613
GetOverlappedResult must be called blocking, waiting for the overlapped
operation to complete.
Cygwin: pty: Add FreeConsole to destructor of pty slave.
- When pseudo console is closed, all the processes attched to the
pseudo console are terminated. This causes the problem reported
in https://sourceware.org/pipermail/cygwin/2020-March/244046.html.
This patch fixes the issue.
Fabian Schriever [Wed, 11 Mar 2020 09:58:05 +0000 (10:58 +0100)]
Fix truncf for sNaN input
Make line 47 in sf_trunc.c reachable. While converting the double
precision function trunc to the single precision version truncf an error
was introduced into the special case. This special case is meant to
catch both NaNs and infinities, however qNaNs and infinities work just
fine with the simple return of x (line 51). The only error occurs for
sNaNs where the same sNaN is returned and no invalid exception is
raised.
Fabian Schriever [Tue, 10 Mar 2020 10:24:12 +0000 (11:24 +0100)]
Fix error in fdim/f for infinities
The comparison c == FP_INFINITE causes the function to return +inf as it
expects x = +inf to always be larger than y. This shortcut causes
several issues as it also returns +inf for the following cases:
- fdim(+inf, +inf), expected (as per C99): +0.0
- fdim(-inf, any non NaN), expected: +0.0
I don't see a reason to keep the comparison as all the infinity cases
return the correct result using just the ternary operation.
While testing the exp function we noticed some errors at the specified
magnitude. Within this range the exp function returns the input value +1
as an output. We chose to run a test of 1m exponentially spaced values
in the ranges [-2^-27,-2^-32] and [2^-32,2^-27] which showed 7603 and
3912 results with an error of >=0.5 ULP (compared with MPFR in 128 bit)
with the highest being 0.56 ULP and 0.53 ULP.
It's easy to fix by changing the magnitude at which the input value +1
is returned from <2^-28 to <2^-32 and using the polynomial instead. This
reduces the number of results with an error of >=0.5 ULP to 485 and 479
in above tests, all of which are exactly 0.5 ULP.
As we were already checking on exp we also took a look at expf. For expf
the magnitude where the input value +1 is returned can be increased from
<2^-28 to <2^-23 without accuracy loss for a slight performance
improvement. To ensure this was the correct value we tested all values
in the ranges [-2^-17,-2^-28] and [2^-28,2^-17] (~92.3m values each).
Do not bother passing optional argument to WriteConsoleA.
Passing a pointer to a local variable to WriteConsoleA is
not actually needed if we're not going to do anything with
what WriteConsoleA would put in there.
For the wpbuf class the pointer argument was made optional,
so it can be just left out; other call places now pass a
NULL pointer instead. The local variables `wn' and `n'
are no unused, so they go away.
The single-precision trigonometric functions show rather high errors in
specific ranges starting at about 30000 radians. For example the sinf
procedure produces an error of 7626.55 ULP with the input
5.195880078125e+04 (0x474AF6CD) (compared with MPFR in 128bit
precision). For the test we used 100k values evenly spaced in the range
of [30k, 70k]. The issues are periodic at higher ranges.
This error was introduced when the double precision range reduction was
first converted to float. The shift by 8 bits always returns 0 as iq is
never higher than 255.
The fix reduces the error of the example above to 0.45 ULP, highest
error within the test set fell to 1.31 ULP, which is not perfect, but
still a significant improvement. Testing other previously erroneous
ranges no longer show particularly large accuracy errors.
Takashi Yano [Mon, 2 Mar 2020 01:12:57 +0000 (10:12 +0900)]
Cygwin: console: Add a workaround for "ESC 7" and "ESC 8".
- In xterm compatible mode, "ESC 7" and "ESC 8" do not work properly
in the senario:
1) Execute /bin/ls /bin to fill screen.
2) Sned CSI?1049h to alternate screen.
3) Reduce window size.
4) Send CSI?1049l to resume screen.
5) Send "ESC 7" and "ESC 8".
After sending "ESC 8", the cursor goes to incorrect position. This
patch adds a workaround for this issue.
Takashi Yano [Mon, 2 Mar 2020 01:12:55 +0000 (10:12 +0900)]
Cygwin: console: Fix setting/unsetting xterm mode for input.
- This patch fixes the issue that xterm compatible mode for input
is not correctly set/unset in some situation such as:
1) cat is stopped by ctrl-c.
2) The window size is changed in less.
In case 1), request_xterm_mode_input(true) is called in read(),
however, cat is stopped without request_xterm_mode_input(false).
In case 2), less uses longjmp in signal handler, therefore,
corresponding request_xterm_mode_input(false) is not called if
the SIGWINCH signal is sent within read(). With this patch,
InterlockedExchange() is used instead of InterlockedIncrement/
Decrement().
Cygwin: ioctl: TIOCINQ: always return number of chars in the inbound queue
So far ioctl(TIOCINQ) could end up returning -1 with errno set to EINVAL
if a non-zero device error mask has been returned by ClearCommError.
This doesn't reflect Linux behaviour, which always returns the number of
chars in the inbound queue, independent of any I/O error condition.
EINVAL was a pretty weird error code to use in this scenario, too.
Fix this by dropping all checking for device errors in the TIOCINQ
case. Just return the number of chars in the inbound queue.
arm: Finish moving newlib to unified syntax for Thumb1
Most code in newlib already uses unified syntax, but just a couple of
laggards remain. This patch removes these and means the the entire
code base has now been converted.
Takashi Yano [Thu, 27 Feb 2020 02:33:50 +0000 (11:33 +0900)]
Cygwin: console: Adjust the detailed behaviour of ESC sequences.
- This patch makes some detailed behaviour of ESC sequences such as
"CSI Ps L" (IL), "CSI Ps M" (DL) and "ESC M" (RI) in xterm mode
match with real xterm.
Corinna Vinschen [Fri, 28 Feb 2020 12:21:05 +0000 (13:21 +0100)]
Cygwin: AF_UNIX: rework fixup_after_exec
fhandler_socket_unix::fixup_after_exec incorrectly calls
fhandler_socket_unix::fixup_after_fork with a NULL parent process
handle. Not only that calling DuplicateHandle with a NULL parent
handle fails, but it's utterly wrong trying to duplicate the handles
at all here.
Rather just set some important values to NULL and reopen the shared
memory region. Create a fixup_helper method to call common code from
fixup_after_fork and fixup_after_exec.
Add comments to other invocations of fixup_after_fork with NULL
handle to mark them as correct this way.
Corinna Vinschen [Fri, 28 Feb 2020 13:31:56 +0000 (14:31 +0100)]
Cygwin: 32 bit: remove old code to 16 bit align stack
Aligning the stack pointer using an asm statement isn't any longer
supported. gcc-9.2.0 generates the following warning:
init.cc:33:46: error: listing the stack pointer register '%esp'
in a clobber list is deprecated [-Werror=deprecated]
[...]
init.cc:33:46: note: the value of the stack pointer after an
'asm' statement must be the same as it was before the statement
Replace the asm expression with the gcc function attribute
`force_align_arg_pointer'. This aligns the stack exactly as
required.
Corinna Vinschen [Fri, 28 Feb 2020 11:39:41 +0000 (12:39 +0100)]
Cygwin: AF_UNIX: fix creating shared mem region in dup
reopen_shmem is accidentally called on the parent fhandler
rather than the child fhandler, and it's called too early.
Make sure to call it on the child and only after its shmem_handle
is valid.
Takashi Yano [Wed, 26 Feb 2020 15:33:01 +0000 (00:33 +0900)]
Cygwin: console: Add support for REP escape sequence to xterm mode.
- In Win10 upto 1809, xterm compatible mode does not have REP
escape sequence which terminfo declares. This patch adds support
for "CSI Ps b" (REP). With this patch, bvi (binary editor) works
normally in Win10 1809. Also, xterm compatible mode does not have
"CSI Pm `" (HPA), "CSI Pm a" (HPR) and "CSI Ps e" (VPR). However,
they do not appear to be declared by terminfo. Therefore, these
have been pending.
Takashi Yano [Wed, 26 Feb 2020 15:32:59 +0000 (00:32 +0900)]
Cygwin: console: Add workaround for broken IL/DL in xterm mode.
- Cygwin console with xterm compatible mode causes problem reported
in https://www.cygwin.com/ml/cygwin-patches/2020-q1/msg00212.html
if background/foreground colors are set to gray/black respectively
in Win10 1903/1909. This is caused by "CSI Ps L" (IL), "CSI Ps M"
(DL) and "ESC M" (RI) control sequences which are broken. This
patch adds a workaround for the issue.
Corinna Vinschen [Wed, 26 Feb 2020 20:08:51 +0000 (21:08 +0100)]
Cygwin: ps: fix compiler warning in ttynam
The helper function ttynam creates a tty name by using sprintf wrongly
on a pretty short buffer. The foramt string only specifies a minimum
field length, not a maximum field length, so gcc-9.2.0 complains:
ps.cc:101:23: warning: 'sprintf' may write a terminating nul past the
end of the destination [-Wformat-overflow=]
Fix this thoroughly by specifying a maximum field width as well as by
using snprintf with a fixed buffer length. Also, drop using a static
buffer in favor of using a buffer in the caller.
Corinna Vinschen [Wed, 26 Feb 2020 19:52:55 +0000 (20:52 +0100)]
Cygwin: cygserver: drop useless packed attribute
...from structs used for data exchange between clients and cygserver.
All of the structs have the same size and member offsets, packed or
unpacked. Keeping the packed attribute results in ominous warnings
from gcc-9.2.0:
cygserver.cc:259:10: warning: taking address of packed member of
'client_request_attach_tty::request_attach_tty' may result in an
unaligned pointer value [-Waddress-of-packed-member]
Corinna Vinschen [Wed, 26 Feb 2020 15:50:34 +0000 (16:50 +0100)]
Cygwin: posix timers: fix uninitialized variable
The variable returning the overrun count from the tracker object after
disarming the overrun counter was not correctly initialized. For some
reason this has only been noticed by gcc-9.2.0, not by the formerly used
gcc-7.4.0.
This problem should not have had any runtime impact. The method
timer_tracker::disarm_overrun_event is supposed to be called in
lock-step with timer_tracker::arm_overrun_event, which in turn
results in the variable getting a valid value.
linux 4.6 x86/cpu: Add advanced power management bits
Bit 11 of CPUID 8000_0007 edx is processor feedback interface.
Bit 12 of CPUID 8000_0007 edx is accumulated power.
Print proper names in /proc/cpuinfo
[missed enabling this 2016 change during previous major cpuinfo update
as no power related changes were made to the Linux files since then]
Joel Sherrill [Tue, 25 Feb 2020 13:38:33 +0000 (07:38 -0600)]
x86_64/i386 fenv: Replace symlink with include fenv_stub.c
Having symlinks for these files led to an issue reported to the RTEMS
Project that showed up using some tar for native Windows to unpack the
newlib sources. It creates symlinks in the tar file as copies of the
files the symlinks point to. If the links appear in the tar file before
the source exists, it cannot copy the file.
The solution in this patch is to convert the files that are symbolic
links into simple files which include the file they were linked to.
This should be more portable and avoids the symbolinc link problem.
Takashi Yano [Mon, 24 Feb 2020 16:12:16 +0000 (01:12 +0900)]
Cygwin: console: Fix segfault on shared_console_info access.
- Accessing shared_console_info before initialization causes access
violation because it is a NULL pointer. The cause of the problem
reported in https://cygwin.com/ml/cygwin/2020-02/msg00197.html is
this NULL pointer access in request_xterm_mode_output() when it is
called from close(). This patch makes sure that shared_console_info
is not NULL before calling request_xterm_mode_output().
Brian Inglis [Fri, 21 Feb 2020 21:28:08 +0000 (14:28 -0700)]
fhandler_proc/cpuinfo: support fast short REP MOVSB
Added in Linux 5.6:
Check FSRM and use REP MOVSB for short copies on systems that have it.
>From the Intel Optimization Reference Manual:
3.7.6.1 Fast Short REP MOVSB
Beginning with processors based on Ice Lake Client microarchitecture,
REP MOVSB performance is enhanced with string lengths up to 128 bytes.
Support for fast-short REP MOVSB is indicated by the CPUID feature flag:
CPUID [EAX=7H, ECX=0H).EDX.FAST_SHORT_REP_MOVSB[bit 4] = 1.
There is no change in the REP STOS performance.
Corinna Vinschen [Thu, 20 Feb 2020 13:48:03 +0000 (14:48 +0100)]
Cygwin: console: ignore NUL byte on write in xterm emulation mode as well
A NUL byte in the output stream got accidentally not handled as IGN char
in xterm console mode. The internal mbtowc conversion doesn't handle
embedded NUL values gracefully, it always stops converting at NUL bytes.
This broke the output of strings with embedded NUL bytes.
Fix this by always skipping IGN chars in the "normal char output loop"
and make sure not to move the cursor one position to the right, as in
legacy console mode.
Thomas Wolff [Sun, 16 Feb 2020 23:00:00 +0000 (00:00 +0100)]
Locale modifier "@cjksingle" to enforce single-width CJK width.
This option follows a proposal in the Terminals Working Group Specifications
(https://gitlab.freedesktop.org/terminal-wg/specifications/issues/9#note_406682).
It makes locale width consistent with the corresponding mintty feature.
Takashi Yano [Mon, 17 Feb 2020 10:29:07 +0000 (19:29 +0900)]
Cygwin: console: Change timing of set/unset xterm compatible mode.
- If two cygwin programs are executed simultaneousley with pipes
in cmd.exe, xterm compatible mode is accidentally disabled by
the process which ends first. After that, escape sequences are
not handled correctly in the other app. This is the problem 2
reported in https://cygwin.com/ml/cygwin/2020-02/msg00116.html.
This patch fixes the issue. This patch also fixes the problem 3.
For these issues, the timing of setting and unsetting xterm
compatible mode is changed. For read, xterm compatible mode is
enabled only within read() or select() functions. For write, it
is enabled every time write() is called, and restored on close().
Nicolas Brunie [Fri, 14 Feb 2020 09:12:25 +0000 (10:12 +0100)]
pow: fix pow(-1.0, NaN)
I think I may have encountered a bug in the implementation of pow:
pow(-1.0, NaN) returns 1.0 when it should return NaN.
Because ix is used to check input vs 1.0 rather than hx, -1.0 is
mistaken for 1.0
Takashi Yano [Mon, 10 Feb 2020 11:42:45 +0000 (20:42 +0900)]
Cygwin: pty: Prevent potential errno overwriting.
- In push_to_pcon_screenbuffer(), open() and ioctl() are called.
Since push_to_pcon_screenbuffer() is called in read() and write(),
errno which is set in read() and write() code may be overwritten
in open() or ioctl() call. This patch prevent this situation.
Takashi Yano [Sun, 9 Feb 2020 14:47:30 +0000 (23:47 +0900)]
Cygwin: pty: Fix state mismatch caused in mintty.
- PTY has a bug reported in:
https://cygwin.com/ml/cygwin/2020-02/msg00067.html.
This is the result of state mismatch between real pseudo console
attaching state and state variable. This patch fixes the issue.
Takashi Yano [Sun, 9 Feb 2020 14:46:59 +0000 (23:46 +0900)]
Cygwin: pty: Inherit typeahead data between two input pipes.
- PTY has a problem that the key input, which is typed during
windows native app is running, disappear when it returns to shell.
(Problem 3 in https://cygwin.com/ml/cygwin/2020-02/msg00007.html)
This is beacuse pty has two input pipes, one is for cygwin apps
and the other one is for native windows apps. The key input during
windows native program is running is sent to the second input pipe
while cygwin shell reads input from the first input pipe.
This patch realize transfering input data between these two pipes.
Takashi Yano [Sun, 9 Feb 2020 14:46:02 +0000 (23:46 +0900)]
Cygwin: pty: Remove debug codes and organize related codes.
- Debug codes used in the early stage of pseudo console support are
removed. (Regarding ALWAYS_USE_PCON and USE_API_HOOK) Along with
this, the codes related to this change are organized.
Takashi Yano [Thu, 6 Feb 2020 10:48:17 +0000 (19:48 +0900)]
Cygwin: pty: Use pinfo() rather than kill() with signal 0.
- PTY code has a problem that tcsh is terminated if the following
command is executed.
true; chcp &
This seems to be caused by invalid pointer access which occurs
when the process exits during the kill() code is execuetd. This
patch avoids the issue by not using kill().
libgloss: Fix lseek semihosting bug on nios2 and m68k
When off_t is 32 bits, the value needs to be sign-extended to 64 bits
before shifting right to extract the high-order word. Previously
negative offsets were incorrectly encoded.
Takashi Yano [Sat, 1 Feb 2020 04:28:39 +0000 (13:28 +0900)]
Cygwin: console: Revise color setting codes in legacy console mode.
- With this patch, foreground color and background color are allowed
to be set simultaneously by 24 bit color escape sequence such as
ESC[38;2;0;0;255;48;2;128;128;0m in legacy console mode.
Ken Brown [Thu, 30 Jan 2020 15:08:21 +0000 (10:08 -0500)]
Cygwin: fstat_helper: always use handle in call to get_file_attribute
When fhandler_base::fstat_helper is called, the handle h returned by
get_stat_handle() should be pc.handle() and should be safe to use for
getting the file information. Previously, the call to
get_file_attribute() for FIFOs set the first argument to NULL instead
of h, thereby forcing the file to be opened for fetching the security
descriptor in get_file_sd().
Jim Wilson [Wed, 29 Jan 2020 22:46:17 +0000 (14:46 -0800)]
RISC-V: Use newlib nano specific libm.
The libm gamma functions use the _gamma_signgam field of the reentrant
structure, which changes offset with the --enable-newlib-reent-small
configure option, which means we need to use a newlib nano specific
version of libm in addition to libc in the nano.specs file. Reported
by Keith Packard. There is a riscv-gnu-toolchain patch that goes
along with this to create the new libm_nano.a file.
Ken Brown [Sat, 25 Jan 2020 18:08:00 +0000 (13:08 -0500)]
Cygwin: AF_LOCAL: fix fcntl and dup if O_PATH is set
Make fhandler_socket_local::dup and fhandler_socket_local::fcntl (a
new method) call fhandler_base::dup and fhandler_base::fcntl if O_PATH
is set.
We're viewing the socket as a disk file here, but there's no need to
implement the actions of fhandler_disk_file::dup and
fhandler_disk_file::fcntl, which do nothing useful in this case beyond
what the fhandler_base methods do. (The extra actions are only useful
when I/O is going to be done on the file.)
Ken Brown [Sat, 25 Jan 2020 12:45:10 +0000 (07:45 -0500)]
Cygwin: AF_LOCAL::fstatvfs: use our handle if O_PATH is set
If O_PATH is set, then the fhandler_socket_local object has a handle
that can be used for getting the statvfs information. Use it by
calling fhandler_base::fstatvfs_by_handle. Without this change,
fhandler_disk_file::fstatfvs would be called on a new fhandler_disk
object, which would then have to be opened.
Ken Brown [Thu, 23 Jan 2020 20:11:15 +0000 (15:11 -0500)]
Cygwin: AF_LOCAL: set appropriate errno on system calls
If an AF_LOCAL socket is opened with O_PATH, all socket system calls
that take a file descriptor argument fail on the resulting descriptor.
Make sure that errno is set as on Linux for those calls that are
implemented on Linux. In almost all cases it is ENOTSOCK. There are
two exceptions:
- sockatatmark(3); errno is EBADF.
- bindresvport(3); errno is EAFNOSUPPORT if the second argument sin
(of type struct sockaddr_in *) is non-NULL and satisfies
sin->sin_family == AF_INET.
Finally, there are two BSD socket system calls implemented on Cygwin
but not Linux: getpeereid(3) and bindresvport_sa(3). Set errno to
ENOTSOCK for these for consistency with the majority of the other
calls.
Ken Brown [Thu, 23 Jan 2020 19:39:15 +0000 (14:39 -0500)]
Cygwin: AF_LOCAL: allow opening with the O_PATH flag
If that flag is not set, or if an attempt is made to open a different
type of socket, the errno is now EOPNOTSUPP instead of ENXIO. This is
consistent with POSIX, starting with the 2016 edition. Earlier
editions were silent on this issue.
Opening is done in a (new) fhandler_socket_local::open method by
calling fhandler_base::open_fs.
Also add a corresponding fhandler_socket_local::close method.
This breaks gnulib's autoconf test for POSIX compatibility of
fflush/fseek. After fflush/fseek, ftello and lseek are out of
sync, with lseek having the wrong offset. This breaks backward
compatibility with Cygwin applications.
Ken Brown [Fri, 24 Jan 2020 20:09:03 +0000 (15:09 -0500)]
Cygwin: FIFO: fstatvfs: use our handle if O_PATH is set
If O_PATH is set, then the fhandler_fifo object has a handle that can
be used for getting the statvfs information. Use it by calling
fhandler_base::fstatvfs_by_handle. Before this change,
fhandler_disk_file::fstatfvs was called on a new fhandler_disk_file
object, which would then have to be opened.
Ken Brown [Fri, 24 Jan 2020 21:02:57 +0000 (16:02 -0500)]
Cygwin: fhandler_disk_file::fstatvfs: refactor
Define a new method fhandler_base::fstatvfs_by_handle, extracted from
fhandler_disk_file::fstatvfs, which gets the statvfs information when
a handle is available.
This will be used in future commits for special files that have been
opened with O_PATH.
Ken Brown [Fri, 24 Jan 2020 19:54:47 +0000 (14:54 -0500)]
Cygwin: fhandler_base::fstat_fs: accomodate the O_PATH flag
Treat a special file opened with O_PATH the same as a regular file,
i.e., use its handle to get the stat information.
Before this change, fstat_fs opened the file a second time, with the
wrong flags and without closing the existing handle. A side effect
was to change the openflags of the file, possibly causing further
system calls to fail.
Currently this change only affects FIFOs, but it will affect
AF_LOCAL/AF_UNIX sockets too once they support O_PATH.