With the previous commit 9ddd48ee1b ("Cygwin: /proc/<PID>/maps:
print real shared region addresses"), the real addresses of
the standard shared regions (cygwin, user, myself, shared console)
are read from the printed process itself. We don't need fixed
addresses anymore, so drop the definitions and simplify open_shared.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
So far, the addresses printed for the shared regions of a process
were faked. The assumption was that the shared regions are always
in the same place in all processes, so we just printed the addresses
of the current process. This is no safe bet. The only safe bet is
the address of the cygheap. So keep track of the addresses in the
cygheap and read the addresses from the cygheap of the observed
processes. Add output for the shared console.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
It's not a good idea to enable high-entropy VA for tools loading the
Cygwin DLL dynamically. The addresses used by HEVA tend to collide with
fixed address areas managed by Cygwin.
Fixes: 60675f1a7e ("Cygwin: decouple shared mem regions from Cygwin DLL")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
For fixed regions (cygwin/user/myself/shared console), try fixed
address first. Fallback to non-fixed region. Don't even try fixed
address if the Cygwin DLL gets dynamically loaded.
For non-fixed regions, try to allocate in a loop within the area
from SHARED_REGIONS_ADDRESS_LOW to SHARED_REGIONS_ADDRESS_HIGH.
Fixes: 60675f1a7e ("Cygwin: decouple shared mem regions from Cygwin DLL")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Fix comments accordingly.
This is in preparation for a change in open_shared, handling shared
regions more cleanly.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
For ages, open_shared uses the shared_locations parameter as
output to indicate if the mapping for a shared region has been
created or just opened. Split this into two parameters. Use
the shared_locations parameter as input only, return the creation
state of the mapping in a bool reference parameter.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
The number of args multiplied by 4 was only required for
32 bit x86 to construct the symbol name correctly. Drop it.
In the assembler code, split the numerical values for "notimp"
and "err" into half words for cleaner layout.
Fix description accordingly.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
vmstat from proc-ps-4.0.x prints "Unable to create system stat structure"
if the /proc/cpuinfo output fails to contain topology info. While
Linux always prints topology info if the kernel has been built with
CONFIG_SMP, Cygwin only prints topology info if the CPU is known to
be multi-core (i. e., the HT feature flag is set).
Fix that by printing topology info all the time, even for single-core
CPUs.
Fixes: e0d48debed ("Fix /proc/cpuinfo topology and cache size info")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
- add '-' option
- make group argument optional
- drop ability to take a numerical group argument
- simplify usage output to bare minimum
- Add manpage and documentation
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
This tool allows to change the primary group for a child process.
The new primary group MUST be part of the supplementary group list
of newgrp's user token.
The command started as child process is specified on the command line.
If it's missing, start the user's default shell with the new primary
group.
TODO: Implement '-' option.
Add command description to documentation.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Default mutex type is PTHREAD_MUTEX_NORMAL.
Attempting to unlock an unowned mutex of that type is specified as
undefined behaviour, not returning EPERM.
mutex7e verfies that attempting to unlock an unowned mutex of type
PTHREAD_MUTEX_ERRORCHECK returns EPERM.
Build all the testcase executables directly using automake, rather than
passing the compiler information into DejaGnu to have it build them.
(This means you get build avoidance for these executables, so they only
get built once, rather than every time you run the test, and makes it
much easier to run them in isolatation against the installed Cygwin,
which is really nice to have when trying to fix broken tests...)
Rename the 'cygrun' subdirectory to 'mingw', and build all the testsuite
MinGW executables there.
Drop sample-miscompile.c (testing that compile failure is detected is
perhaps useful, but not here...)
objdir isn't a predefined output variable in Automake (any more?), so
this was just using the absolute path /testsuite as the test's temporary
directory. Use builddir instead.
So far the capability bits were stored in the .cygwin_dll_common
R/W section because we overwrite the is_server bit. Just don't.
Move the bit to class wincapc instead and define all wincaps
bitfields const.
Fixes: 8937c103ed ("* wincap.cc (all wincaps): Store in .cygwin_dll_common section same as wincap. Add comment to explain why.")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
_pinfo::ctty has two special values other than the device id of
the allocated ctty:
-1: CTTY is not initialized yet. Can be associated with the TTY
which is associated with the session leader.
-2: CTTY has been released by setsid(). Can be associate only with
new TTY which is not associated with any other session as CTTY,
but cannot be associate with the TTYs already associated with
other sessions.
This patch adds the comments in some source files.
Suggested-by: Corinna Vinschen <corinna@vinschen.de>
Signedoff-by: Takashi Yano <takashi.yano@nifty.ne.jp>
The commit 25c4ad6ea5 did not fix the CTTY behavior enough. For
example, in the following test case, TTY will be associated as
a CTTY on the second open() call even though the TTY is already
CTTY of another session. This patch fixes the issue.
#include <unistd.h>
#include <sys/fcntl.h>
int main()
{
if (fork () == 0) {
char *tty = ttyname(0);
int fd;
setsid();
fd = open(tty, O_RDWR);
close(fd);
fd = open(tty, O_RDWR);
usleep (60000000L);
}
return 0;
}
Fixes: 25c4ad6ea5 ("Cygwin: pinfo: Align CTTY behavior to the
statement of POSIX.")
Reviewed-by: Corinna Vinschen <corinna@vinschen.de>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
After the commit 30add3e6b3, the problem:
https://cygwin.com/pipermail/cygwin/2022-December/252759.html
occurs rarely. It seems that myself_pinfo should be initialized
where the timing before child_copy() and after cygheap allocation.
This patch moves the initialization there.
Fixes: 30add3e6b3 ("Cygwin: exec: don't access cygheap before it's
initialized")
Reported-by: Brian Inglis <Brian.Inglis@Shaw.ca>
Reviewed-by: Corinna Vinschen <corinna@vinschen.de>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
The code computing the mask of pending signals used the per-queued
signal TLS pointer without checking it for NULL. Fix this by using
the process-wide signal mask in that case.
Fixes: 195169186b ("Cygwin: wait_sig: allow to compute process-wide mask of pending signals")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Since 4e7817498e, we're just running the tests against the installed
DLL. We're arranging to put the build directory on the path, but since
it doesn't contain cygwin1.dll (since it's built with a different name
and renamed on installation), that doesn't have any effect.
Arrange to place the just-built DLL into a directory which the testsuite
can place on it's path (while running the test, but not while compiling
it).
Also fix any remaining references to cygwin0.dll in testsuite,
documentation and comments.
Fixes: 4e7817498e ("Cygwin: Makefile: Drop all the "test dll" considerations")
The code generating the signal info in _pinfo::siginfo() and in
commune_process() are doing the same thing. Create a local static
function commune_process_siginfo() to have the code in one place
only. Remove a useless sigpending() call.
Fixes: 9a3c058f66 ("Cygwin: /proc/<PID>/status: Fill SigPnd, SigBlk and SigIgn values with life")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Previously, the console device could not be accessed from other terminals.
Due to this limitation, GNU screen and tmux cannot be opened in console.
With this patch, console device can be accessed from other TTYs, such as
other consoles or ptys. Thanks to this patch, screen and tmux get working
in console.
Reviewed-by: Corinna Vinschen <corinna@vinschen.de>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
The devices /dev/conin,conout,console were wrongly visible from ptys,
though they are inaccessible. This is because fhandler_console::exists()
returns true due to existing invisible console. This patch makes these
devices invisible from ptys.
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
POSIX states "A terminal may be the controlling terminal for at most
one session."
https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap11.html
However, in cygwin, multiple sessions could be associated with the
same TTY. This patch aligns CTTY behavior to the statement of POSIX.
Reviewed-by: Corinna Vinschen <corinna@vinschen.de>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
https://cygwin.com/pipermail/cygwin/2022-December/252737.html
If the less is started from non-cygwin shell and window size is
changed, it will hang-up when quitting. The cause of the proglem is
that less uses longjump() in signal handler. If the signal handler
is called while cygwin is acquiring the mutex, cygwin loses the
chance to release mutex. With this patch, the mutex is released
just before calling kill_pgrp() and re-acquired when kill_pgrp()
returns.
Reported-by: Gregory Mason <grmason@epic.com>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
After 90236c3a2c, the testsuite is failing, as the cygwin0.dll
referenced by the implib that testsuite programs are linked with doesn't
exist anymore.
We don't need to make and link the testsuite with a specially named DLL,
as the cygwin DLL (since 526b0fbca3) takes into consideration the path
it's executing from to define separate "Cygwin installations", which
don't interact.
Fixes: 90236c3a2c ("Cygwin: Makefile: build new-cygwin1.dll in a single step")
Commit c1023ee353 introduced a split between mount flags and
path flags. It didn't initialize symlink_info::path_flags in
path_conv::check, because that's done in symlink_info::check.
However, there are two code paths expecting symlink_info::path_flags
being already initialized and both skip symlink_info::check.
Make sure symlink_info::path_flags is initalized to 0 early in
path_conv::check.
Fixes: c1023ee353 ("Cygwin: path_conv: decouple path_types from mount types")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Rather than guessing, based on just the presence of libbfd, add an
explicit configuration option, to build dumper or not, defaulting to
building it.
This might have some use when bootstrapping Cygwin for a new
architecture, or when building your own Cygwin-targetted cross-compiler,
rather than installing one from the copr, along with the dependencies of
libbfd.
Tweak slightly to allow implementing entire {w}mem{p}{cpy,move}
family:
Add WIDE macro processing for wmem* and POST macro processing for
memp* functions.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
The complicated build routine was only required because we needed
the .cygheap section at the end of the file, and the debug sections
broke this. Now that the cygheap is out of the way, we really don't
have to do this anymore, and the build can just generate a DLL with
integrated debug info.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Make sure to create a new cygwin_version.c if it either doesn't
exist yet, or if it would be different from the former file.
This avoids unnecessary DLL rebuilding.
Fixes: 97eb64b909 ("Cygwin: uname: generate default release string from git as well")
Fixes: 4949a82cde ("Cygwin: uname: fix building in non-git source dir")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Move the drive substitution code after the call to
GetFinalPathNameByHandleW into a local function revert_virtual_drive
and add code to handle non-remote virtual drives, i. e., those
created with the subst command. (Try to) make sure that virtual
drives are never treated like symlinks.
Fixes: 19d59ce75d ("Cygwin: path_conv: Rework handling native symlinks as inner path components")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
https://cygwin.com/pipermail/cygwin/2022-December/252628.html
After the commit 9e4d308cd5, the performance of read from non-cygwin
pipe has been degraded. This is because select_sem mechanism does not
work for non-cygwin pipe. This patch fixes the issue.
Fixes: 9e4d308cd5 ("Cygwin: pipe: Adopt FILE_SYNCHRONOUS_IO_NONALERT
flag for read pipe.")
Reported-by: tryandbuy <tryandbuy@proton.me>
Reviewed-by: Corinna Vinschen <corinna@vinschen.de>
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
Until Cygwin 3.3.6, we define __LARGE64_FILES unconditionally, so we
were using the type __sFILE64 even for 64 bit. That was lazy and wrong.
so commit 2902b3a09e ("Cygwin: drop requirement to build newlib's
stdio64") tried to fix that.
Unfortunately this patch forgot to take the exposure of the typename
__sFILE64 in userspace into account. This leads to trouble in C++ due
to name mangling.
Commit 0f376ae220 tried to fix this by just renaming __sFILE to
__sFILE64 by using a macro. While __sFILE and __sFILE64 are the same
size, they are not exactly congruent.
To avoid backward compatibility problems, make sure to define FILE
as the real __sFILE64, and make sure that __sFILE is not defined at
all on Cygwin.
Fixes: 0f376ae220 ("Cygwin: rename __sFILE to __sFILE64 for backward
compatibility")
Fixes: 2902b3a09e ("Cygwin: drop requirement to build newlib's stdio64")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Until Cygwin 3.3.6, we define __LARGE64_FILES unconditionally, so we
were using the type __sFILE64 even for 64 bit. That was lazy and wrong.
so commit 2902b3a09e ("Cygwin: drop requirement to build newlib's
stdio64") tried to fix that.
Unfortunately this patch forgot to take the exposure of the typename
__sFILE64 in userspace into account. This leads to trouble in C++ due
to name mangling.
Fix this by redefining __sFILE to __sFILE64. The type name is very much
internal, so it doesn't really matter, except for the fact that it needs
to stay backward compatible so as not to break building against C++ libs
built under older versions of Cygwin.
Fixes: 2902b3a09e ("Cygwin: drop requirement to build newlib's stdio64")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
commit 97eb64b909 broke building outside of a git dir, because
the git describe command would fail. Fix this by checking if
we're in a git tree at all and just generate an empty string
as version string. Use this in uname_x to generate a fallback
version.
Fixes: 97eb64b909 ("Cygwin: uname: generate default release
string from git as well"
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
When building a release with cygport, we get uname version info
from cygport, which in turn gets version info from `git describe'.
During development, the release info for local builds was not
that helpful yet. Fix that, by creating version info from
`git describe' if CYGPORT_RELEASE_INFO isn't given. Make sure to
always force rebuild of the version info to pick up source file
changes as well as git actions.
Rearrange code slightly to generate machine info first, release info
after that. Use snprintf to generate release string safely.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Commit 188d5f6c9a erroneously moved the mcountFunc.S file to the
TARGET_FILES target, rather than keeping it in GMON_FILES. The
result is that the __fentry__ entry point is now entirely undefined,
so `gcc -pg' is broken.
Create new target-specific GMON_TARGET_FILES and move mcountFunc.S
into it. Add $(GMON_TARGET_FILES) to GMON_FILES.
Fixes: 188d5f6c9a ("Cygwin: x86_64: add wmemset assembler entry point")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
DiscardVirtualMemory, PrefetchVirtualMemory and
GetSystemTimePreciseAsFileTime are available since
Windows 8.1. Merge PseudoConsole functions into
kernel32 function block.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Only required for Windows 7.
This in turn allows to drop the helper_pid and related
methods from fhandler_pty_common.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Only required for Windows 7.
This allows to remove fhandler_pipe::get_query_hdl_per_system(),
too.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Using ulinks here makes the result work on cygwin.com only, while
xrefs to FAQs are creating realtive links. Add xreflabel where
appropriate.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
fetch_account_from_windows shortcuts the current user in that
it takes the user's domain SID and just adds the matching RID
from the token's primary group to create a group SID.
How wrong this is can be very simply reproduced:
Assuming you run a native process, like cmd, with primary group
set to the Administrators builtin group. Run Cygwin's id(1) as
child process. id(1) will print a non-existent group as primary
group and also add it to the group list.
This can only be avoided by not special casing the current user
and thus not creating a group SID from partial information.
Fixes: 6cc7c925ce ("(pwdgrp::fetch_account_from_windows): Default primary group for the
current user to primary group from user token.")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
internal_getlogin overwrites the process token primary group if it
differs from the primary group as stored in the passwd DB.
However, this also overwrites the primary group of the process if
it has been deliberately changed by a former process (e. g., newgrp),
and the current process has a non-Cygwin process as parent.
Our docs claim we restrict overwriting the primary group to local,
non-domain user accounts anyway, and it was actually meant this way.
So check for exactly that before overwriting the primary group
in the token: It's only allowed if the user is a local account
and the primary group in the token is still the default group
"None".
Fixes: 6cc7c925ce ("(internal_getlogin): Give primary group
from user token more weight.")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Originally the code was written to allow three ways of prefixing
accounts and to freely define a domain/account separator. This code
has been disabled even before being officially released, and it was
never re-enabled. Given there has been no complaints for eight years
now, drop this code eventually. Just add a macro to define the
domain/account separator statically.
Fixes: cc332c9e27 ("(cygheap_pwdgrp::nss_init_line): Disable db_prefix
and db_separator settings. Add comment")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
This is a long-standing thinko.
When you exec a process, dll_crt0_0 in the child process calls
child_info_spawn::handle_spawn(). handle_spawn() initialises the
cygheap.
Now consider calling strace. Strace is a non-Cygwin process dynamically
loading cygwin1.dll via LoadLibrary. This in turn initializes the DLL:
- dll_crt0_0 finds that the process it attaches to has been exec'd, so
child_info_spawn::handle_spawn() is called.
- If the DLL is being dynamically loaded, handle_spawn() calls
child_info_spawn::get_parent_handle(). This in turn tries to set
the moreinfo->myself_pinfo value inside the cygheap to NULL.
- However, at this time, the cygheap has not yet been initialized. This
only occurs in the cygheap_fixup_in_child() call after get_parent_handle()
returns.
--> SEGV
This thinko never had a negative side effect, because the cygheap was
pre-allocated at DLL load time until commit 2f9b8ff00c ("Cygwin:
decouple cygheap from Cygwin DLL"). With 2f9b8ff00c, the cygheap
actually doesn't exist until after the call to cygheap_fixup_in_child().
Fix this problem by moving the assignment after the call to
cygheap_fixup_in_child().
Fixes: 3de7be4c1d ("* DevNotes: Add entry cgf-000007. [...]")
Fixes: 2f9b8ff00c ("Cygwin: decouple cygheap from Cygwin DLL")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
After decoupling cygheap from the DLL loading address, the check
for a different _cygheap_start has gone. So the matching string
check in multiple_cygwin_problem is obsolete now.
Fixes: 2f9b8ff00c ("Cygwin: decouple cygheap from Cygwin DLL")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
In commit 2dab880c96 I did a mistake when
I copied the new fhandler_serial::switch_modem_lines() from my modified
3.3.6 branch to the current master and I left a copy paste error. This
patch fixes that error.
Fixes: 2dab880c96 ("Cygwin: fix TIOCMBIS/TIOCMBIC not working when using usbser.sys")
- Previously, we have two tty.cc, one is winsup/cygwin/tty.cc and
the other is winsup/cygwin/fhandler/tty.cc. This is somewhat
confusing. This patch renames fhandler/tty.cc to fhandler/pty.cc.
When creating regular Cygwin test releases we need a way to
automate unambiguous version information based on the output
of `git describe'. Allow to inject a release string via a
preprocessor macro CYGPORT__RELEASE_INFO. Change the default
release info to recognize a local, non-distro build.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
We're going to switch to regular test releases, rather than the
old, handcrafted snapshots method.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
- The codes related to pty and console in spawn.cc have been moved
into the new class fhandler_termios::spawn_worker, and make spawn.cc
call them. The functionality has not been changed at all.
In winsup/cygwin/fhandler/serial.cc, the function
fhandler_serial::switch_modem_lines() is called when TIOCMBIS/TIOCMBIC
are used in an ioctl() call.
This function uses EscapeCommFunction() for setting and resetting RTS
and DTR signals of a serial port. Unfortunately, this function does not
work on USB CDC devices.
This is not a true bug of CYGWIN but an issue of the usbser.sys driver,
from Windows 2000 to the latest Windows 11. Both 32bit and 64bit
versions of the operating system are affected. Actually, I tested
EscapeCommFunction() also when using a real UART, based on the
traditional 16550 driver and it works fine. Using thirdy party CDC
drivers, like the one provided by FTDI for their USB bridge chips,
probably also works.
However, it is also possible to drive the RTS/DTR signals by writing
their state with SetCommState(), which proved to be working fine all
types of connection. This is also a better solution for handling these
signals since RTS/DTR can be set at the same time rather than having two
separate calls with a visible delay between them.
Report 0 instead of 268435455 (i.e. 0xFFFFFFF) in the tty field of
/proc/*/stat for processes without a controlling terminal. This is what
the procps utility expects when selecting or excluding such processes.
This adds an extra section to the stackdump, which lists the loaded
modules and their base address. This is perhaps useful as it makes it
immediately clear if RandomCrashInjectedDll.dll is loaded...
Future work: It seems like the 'InMemoryOrder' part of
'InMemoryOrderModuleList' is a lie?
> Loaded modules
> 000100400000 segv-test.exe
> 7FFF2AC30000 ntdll.dll
> 7FFF29050000 KERNEL32.DLL
> 7FFF28800000 KERNELBASE.dll
> 000180040000 cygwin1.dll
> 7FFF28FA0000 advapi32.dll
> 7FFF29F20000 msvcrt.dll
> 7FFF299E0000 sechost.dll
> 7FFF29B30000 RPCRT4.dll
> 7FFF27C10000 CRYPTBASE.DLL
> 7FFF28770000 bcryptPrimitives.dll
This adds an additional column to the stack trace in a .stackdump file,
which gives the stack frame return address as a module name+offset. This
makes it a possible to convert the address to a function name without
having to guess what module the address belongs to.
> Stack trace:
> Frame Function Args
> 0007FFFFCC30 0001004010E9 (000180048055, 000180046FA0, 000000000002, 00018031E160) segv-test.exe+0x10E9
> 0007FFFFCD30 0001800480C1 (000000000000, 000000000000, 000000000000, 000000000000) cygwin1.dll+0x80C1
> 0007FFFFFFF0 000180045C86 (000000000000, 000000000000, 000000000000, 000000000000) cygwin1.dll+0x5C86
> 0007FFFFFFF0 000180045D34 (000000000000, 000000000000, 000000000000, 000000000000) cygwin1.dll+0x5D34
> End of stack trace
Loosely based on this patch [1] by Brian Dessent.
[1] https://cygwin.com/pipermail/cygwin-patches/2008q1/006306.html
Every time the cygheap is initialized, that is, on each fork
or exec, cygheap_init() *again* computes the bucket size values
and stores them in the cgyheap, albeit they are always the
same values anyway.
Make bucket_val a local const array, statically initialized
instead.
Fixes: 61522196c7 ("* Merge in cygwin-64bit-branch.)"
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Now that the cygheap isn't part of the CYgwin DLL anymore, we have a
known memory location which is not known in maps output. Fix that by
checking for cygheap address (same in all processes) and add to output.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Prior to 591fec858d01 ("Cygwin: decouple cygheap from Cygwin DLL")
the .cygheap section was required to stay the last section in the
final binary. That required some juggling with objcopy to make
sure the .gnu_debuglink section is prior to the .cygheap section
in the final DLL.
This isn't required anymore, so just drop it.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Another reason ASLR may fail is the coupling of the standard shared
mem regions (global, userinfo, process info, shared console) to the
address of the Cygwin DLL. They are always placed in fixed addresses
preceeding the Cygwin DLL's address. With ASLR this is bound to fail.
Use a fixed, unused memory area to place the shared mem regions.
This also allows to simplify the shared memory creation. There's
no reason anymore to rebase the regions and rather than offsets,
just use the addresses directly.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
One reason that ASLR is tricky is the fact that the cygheap
is placed at the end of the DLL and especially that it's expected
to be growable. To support ASLR, this construct must go.
Define dedicated cygheap memory region and reserve entire region.
Commit 3 Megs, as was the default size of the cygheap before.
Fix linker script accordingly, drop a now useless version check
in get_cygwin_startup_info().
Collect all info about memory layout in one header file, so
the mem layout is documented in one logical place and not
in heap.cc arbitrarily.
Use info from this file throughout.
This is to prepare for ASLR support.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Exception handling was *still* printing addresses as 44 bit values,
but Windows supports a 48 bit virtual address space since Windows
8.1. Fix that.
Fixes: e1254add73 ("Cygwin: Allow accessing 48 bit address space in Windows 8.1 or later")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
.com is a remnant from the past. There are only five executables
left:
chcp.com
format.com
mode.com
more.com
tree.com
Calling them on the command line already requires to use the
suffix anyway. So drop useless .com test from the execve test
for scripts (they are handled earlier in the same function
as executables) and do not handle them like .exe suffixes in
other functions.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
This reverts commit 7589034cc3.
The previous commit 14816de9af69 ("Cygwin: spawn: drop special handling
for cmd.exe and command.com") make this patch unnecessary. The filename
argument (i. e., run_path in the caller) is now always non-NULL.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Apparently at one point handling cmd.exe and command.com special
made sense, but what that should be has never been documented.
There's also no clear reason why cmd.exe is different from any
other native executable. Additionaly, checking for command.com
is entirely useless on 64 bit Windows anyway.
Just drop this code.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
- If the command executed is 'cmd.exe /c [...]', runpath in spawn.cc
will be NULL. In this case, is_console_app(runpath) check causes
access violation. This case also the command executed is obviously
console app., therefore, treat it as console app to fix this issue.
Addresses: https://github.com/msys2/msys2-runtime/issues/108
According to POSIX and the Linux man page, select(2) is supposed to
report read ready if a file is at EOF. In the case of a FIFO, this
means that the pipe is empty and there are no writers. But there
seems to be an undocumented exception, observed on Linux and other
platforms: If no writer has ever been opened, then select(2) does not
report read ready. This can happen if a reader is opened with
O_NONBLOCK before any writers have opened.
This commit makes Cygwin consistent with those other platforms by
introducing a special EOF test, fhandler_fifo::select_hit_eof, which
returns false if there's never been a writer opened.
To implement this we use a new variable '_writer_opened' in the FIFO's
shared memory, which is set to 1 the first time a writer opens. New
methods writer_opened() and set_writer_opened() are used to test and
set this variable.
Addresses: https://cygwin.com/pipermail/cygwin/2022-September/252223.html
The patch fixing the alignment of _cygtls::context accidentally
pushed the desperate attempt to automate the alignment by using
another, non-working variation of attribute((aligned)). Drop it.
Fixes: dcab768cb9 ("Cygwin: cygtls: fix context alignment")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
A hang was encountered, apparently triggered by commit 63b503916d,
changing tls_pathbufs from malloc'ed to HeapAlloc'ed memory. After
lengthy debugging it transpired that adding the heap handle to the
tls_pathbuf struct added 8 bytes to the cygtls area, thus moving
the "context" member by 8 bytes, too, so it was suddently unaligned.
Fix this for now by changing the alignment.
Fix this once and for all, by adding code to the gentls_offsets script
to check if the alignment of the "context" member is 16 bytes. If not,
print a matching error message, remove the just generated file, and exit
with error.
FIXME: It would be really nice to find a way to auomate the correct
alignment of the "context" member, but I don't see any way to use
alignment attributes to get what we need here.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
while debugging a problem introduced in commit
63b503916d ("Cygwin: tls_pathbuf: Use Windows heap")
a hang in fork was encountered using the original implementation
of tls_pathbuf:
Using tmp_pathbuf inside the code block guarded by __malloc_trylock
may call malloc from tmp_pathbuf::w_get and thus trying to lock an
exclusive SRW lock recursively, which results in a deadlock.
Allocate a small SECURITY_ATTRIBUTES block on the stack rather than
allocating a 64K tmp_pathbuf. This avoids the potential malloc call.
Drop the __malloc_trylock call entirely. There must not be a malloc
call inside the frok::parent block guarded by __malloc_lock, and
just trying to lock is too dangerous inside fork while other threads
might actually chage the content of the heap. Additionally, add a
comment frowning on malloc usage inside tis code block.
Fixes: 44a79a6eca ("Cygwin: convert malloc lock to SRWLOCK")
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Rather than using malloc/free for the buffers, we're now using
HeapAlloc/HeapFree on a HEAP_NO_SERIALIZE heap created for this
thread.
Advantages:
- Less contention. Our malloc/free doesn't scale well in
multithreaded scenarios
- Even faster heap allocation by using a non serialized heap.
- Internal, local, temporary data not cluttering the user heap.
- Internal, local, temporary data not copied over to child process
at fork().
Disadvantage:
- A forked process has to start allocating temporary buffers from
scratch. However, this should be alleviated by the fact that
buffer allocation usually reaches its peak very early in process
runtime, so the longer the proceess runs, the less buffers have
to allocated, and, only few processes don't exec after fork
anyway.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
shmat may call shmget. shmget locks by itself as necessary,
so there's no reason to keep the lock active and recurse into
the lock. Use SRWLOCK and only lock as required.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
So far we use a single muto to guard three different datastructures
inside class authz_ctx: the authz HANDLE, the user context HANDLE
and the context cache list. Split the single muto into three
independent SRWLOCKs and guard all datastrcutures as necessary to
avoid thread contention.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
The old technique was from a time when we had to reduce stack pressure
by moving 64K buffers elsewhere. It was implemented using a static
global buffer, guarded by a muto. However, that adds a lock which may
unnecessarily serialize threads.
Use Windows heap buffers per invocation instead. HeapAlloc/HeapFree are
pretty fast, scale nicely in multithreaded scenarios and don't serialize
threads unnecessarily.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
CloseHandle gets redefined to a macro calling an internal function
in debug.h when building with -DDEBUGGING, but profiler has no access
to that function.
Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
Corinna Vinschen
Jon Turney
Takashi Yano
Brian Inglis
Carlo Bramini
Andy Koppe
Hamish McIntyre-Bhatty
Brian Inglis
Ken Brown