summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Cygwin: FIFO: minor change - use NtCloseKen Brown2020-05-081-16/+16
| | | | | Replace CloseHandle by NtClose since all handles are created by NT functions.
* Revert "localtime define _DIAGASSERT" and followups affecting localtime.ccCorinna Vinschen2020-05-041-302/+236
| | | | | | | | | | | | | | | This reverts commits 453b6d17bf0581e55258c5eb6dc7afbad1927cab, 489a47d6036660a6cbab1addfb2b7678c0de8bf2, 3003c3dacd157c11f5e035bc18bdd30631800720, 9e29639ca01208a4a8bac5b9d0f6491b34b1017e, a40701c7dc9970457a19728e371c71e7501057e7, 0a41de2725dbf2f325e55016d85d76560845ff8d, b8aa5f7a0fe70a9a53fe4019227271a78758ae67, 0f4bda879283eaf013f86dd80e5588721a811ea0, 65bf580752db7ac01ecae19dc56aa26ea34f5e47, 3f0c2ac96e7c493bbfc2bc1b0c1274445b756a9d, 76d4d40b8b23d046473a5946cf6b83faa36ae7be, f2e06d8af5f390d81b64375a69ca834ca19d6029.
* setup_pseudoconsole(): handle missing/incorrect helper gracefullyJohannes Schindelin2020-05-041-1/+17
| | | | | | | | | | | | | When `cygwin-console-helper.exe` is either missing, or corresponds to a different Cygwin runtime, we currently wait forever while setting up access to the pseudo console, even long after the process is gone that was supposed to signal that it set up access to the pseudo console. Let's handle that more gracefully: if the process exited without signaling, we cannot use the pseudo console. In that case, let's just fall back to not using it. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
* localtime define _DIAGASSERTCorinna Vinschen2020-04-291-2/+4
|
* localtime 1.82Corinna Vinschen2020-04-281-22/+15
|
* localtime 1.81Corinna Vinschen2020-04-281-104/+101
|
* localtime 1.80Corinna Vinschen2020-04-281-4/+12
|
* localtime 1.79Corinna Vinschen2020-04-281-3/+3
|
* localtime 1.78Corinna Vinschen2020-04-281-64/+68
|
* localtime 1.77Corinna Vinschen2020-04-281-3/+3
|
* localtime 1.76Corinna Vinschen2020-04-281-3/+3
|
* localtime 1.75Corinna Vinschen2020-04-281-3/+8
|
* localtime 1.74Corinna Vinschen2020-04-281-88/+133
|
* localtime 1.73Corinna Vinschen2020-04-271-7/+17
|
* Cygwin: localtime.cc: reformat for easier patchingCorinna Vinschen2020-04-271-8/+10
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: add release message for fixed floppy bugsCorinna Vinschen2020-04-241-0/+4
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: raw disk I/O: lock floppys as wellCorinna Vinschen2020-04-241-2/+2
| | | | | | | | | The workaround to access the full disk required since Vista and described in http://support.microsoft.com/kb/942448 (NOT ACCESSIBLE at the time of writing this commit message) is required on floppy drives as well. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: raw disk I/O: Fix return value in error caseCorinna Vinschen2020-04-241-3/+4
| | | | | | | | | | | | | | | The cast to generate the return value uses a DWORD variable as test and set value. The error case is the constant -1. Given the type of the other half of the conditioal expression, -1 is cast to DWORD as well. On 64 bit, this results in the error case returning a 32 bit -1 value which is equivalent to (ssize_t) 4294967295 rather than (ssize_t) -1. Add a fixing cast. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: file I/O: make sure to treat write return value as ssize_tCorinna Vinschen2020-04-241-3/+3
| | | | | | | | The return type of fhandler write methods is ssize_t. Don't use an int to store the return value, use ssize_t. Use ptrdiff_t for the buffer size. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: accounts: Report unrecognized db_* nsswitch.conf keywordsDavid Macek via Cygwin-patches2020-04-221-0/+2
| | | | Signed-off-by: David Macek <david.macek.0@gmail.com>
* Cygwin: accounts: Don't keep old schemes when parsing nsswitch.confDavid Macek via Cygwin-patches2020-04-222-4/+8
| | | | | | | | | | | | | | | | | The implicit assumption seemed to be that any subsequent occurence of the same setting in nsswitch.conf is supposed to rewrite the previous ones completely. This was not the case if the third or any further schema was previously defined and the last line defined less than that (but at least 2), for example: ``` db_home: windows cygwin /myhome/%U db_home: cygwin desc ``` Let's document this behavior as well. Signed-off-by: David Macek <david.macek.0@gmail.com>
* Cygwin: accounts: Unify nsswitch.conf db_* defaultsDavid Macek via Cygwin-patches2020-04-222-21/+11
| | | | Signed-off-by: David Macek <david.macek.0@gmail.com>
* Cygwin: Add David Macek to CONTRIBUTORSCorinna Vinschen2020-04-211-0/+1
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: symlinks: fix WSL symlink creation if cygdrive prefix is /Corinna Vinschen2020-04-211-4/+6
| | | | | | | | | | | | | | | | | | If the cygdrive prefix is /, then the following happens right now: $ ln -s /tmp/foo . $ ls -l foo lrwxrwxrwx 1 user group 12 Apr 15 23:44 foo -> /mnt/tmp/foo Fix this by skipping cygdrive prefix conversion to WSL drive prefix "/mnt", if the cygdrive prefix is just "/". There's no satisfying way to do the right thing all the time in this case anyway. For a description and the alternatives, see https://cygwin.com/pipermail/cygwin-developers/2020-April/011859.html Also, fix a typo in a comment. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* cygheap_pwdgrp: Handle invalid db_* entries correctlyDavid Macek via Cygwin-patches2020-04-171-1/+4
| | | | | | | | If the first scheme in db_* was invalid, the code would think there were no schemes specified and replace the second scheme with NSS_SCHEME_DESC. Signed-off-by: David Macek <david.macek.0@gmail.com>
* Cygwin: faq: disable outdated "sshd in domain" faqCorinna Vinschen2020-04-141-0/+2
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* proc_cpuinfo: Add PPIN support for AMDBrian Inglis2020-04-141-0/+1
| | | | | | Newer AMD CPUs support a feature called protected processor identification number (PPIN). This feature can be detected via CPUID_Fn80000008_EBX[23].
* Cygwin: utils: override definition of PMEM_EXTENDED_PARAMETERCorinna Vinschen2020-04-072-0/+2
| | | | | | | | | | PMEM_EXTENDED_PARAMETER is defined in the local winlean.h as long as mingw-w64 doesn't define it (in winnt.h). ntdll.h needs the definition for declaring NtMapViewOfSectionEx. cygpath.cc and ps.cc both include ntdll.h but not winlean.h, so they complain about the missing definition. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: mmap_alloc: fix comment to document using the extended memory APICorinna Vinschen2020-04-071-10/+13
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: threads: use extended memory API if availableCorinna Vinschen2020-04-071-3/+48
| | | | | | | | | | | | | So far Cygwin was jumping through hoops to restrict memory allocation to specific regions. With the advent of VirtualAlloc2 and MapViewOfFile3 (and it's NT counterpart NtMapViewOfSectionEx), we can skip searching for free space in the specific regions and just call these functions and let the OS do the job more efficiently and less racy. Use VirtualAlloc2 on W10 1803 and later in thread stack allocation. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: mmap: use extended memory API if availableCorinna Vinschen2020-04-071-10/+74
| | | | | | | | | | | | | So far Cygwin was jumping through hoops to restrict memory allocation to specific regions. With the advent of VirtualAlloc2 and MapViewOfFile3 (and it's NT counterpart NtMapViewOfSectionEx), we can skip searching for free space in the specific regions and just call these functions and let the OS do the job more efficiently and less racy. Use the new functions on W10 1803 and later in mmap. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: memory: declare extended memory APICorinna Vinschen2020-04-075-0/+69
| | | | | | | | | | | | | Windows 10 1803 introduced an extended memory API allowing to specify memory regions allocations are to be taken off. In preparation of using this API, define the struct MEM_EXTENDED_PARAMETER and friends. Declare and allow to autoload the functions VirtualAlloc2 and NtMapViewOfSectionEx. Introduce a wincap flag has_extended_mem_api. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: threads: add missing include of mmap_alloc.hCorinna Vinschen2020-04-071-0/+1
| | | | | | This is needed for using mmap_alloc.alloc() Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: threads: lower thread size from pool to 256 MegsCorinna Vinschen2020-04-071-1/+1
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: threads: use mmap area to fulfill requests for big stacksCorinna Vinschen2020-04-071-4/+6
| | | | | | | Otherwise big stacks have a higher probability to collide with randomized PEBs and TEBs after fork. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: symlinks: Allow traversing WSL symlinksCorinna Vinschen2020-04-051-0/+6
| | | | | | | | | | | | Unfortunately Windows doesn't understand WSL symlinks, despite being a really easy job. NT functions trying to access paths traversing WSL symlinks return the status code STATUS_IO_REPARSE_TAG_NOT_HANDLED. Handle this status code same as STATUS_OBJECT_PATH_NOT_FOUND in symlink_info::check to align behaviour to traversing paths with other non-NTFS type symlinks. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: docs: revamp docs explaining symlinksCorinna Vinschen2020-04-034-77/+146
| | | | | | | The descriptions of symlink handling are a bit dated, so revamp them and add the new WSL symlink type. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: symlinks: create WSL symlinks on supporting filesystemsCorinna Vinschen2020-04-033-26/+105
| | | | | | | | | WSL symlinks are reparse points containing a POSIX path in UTF-8. On filesystems supporting reparse points, use this symlink type. On other filesystems, or in case of error, fall back to the good old plain SYSTEM file. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: symlinks: fix WSL symlinks pointing to /mntCorinna Vinschen2020-04-021-15/+35
| | | | | | | | | | Commit 4a36897af392 allowed to convert /mnt/<drive> path prefixes to Cygwin cygdrive prefixes on the fly. However, the patch neglected WSL symlinks pointing to the /mnt directory. Rearrange path conversion so /mnt is converted to the cygdrive prefix path itself. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* winsup/cygwin: remove defines added in mingw-w64 v7.0.0Biswapriyo Nath2020-04-013-17/+0
| | | | Signed-off-by: Biswapriyo Nath <nathbappai@gmail.com>
* Cygwin: symlinks: support WSL symlinksCorinna Vinschen2020-03-273-0/+83
| | | | | | | | Treat WSL symlinks just like other symlinks. Convert absolute paths pointing to Windows drives via /mnt/<driveletter> to Windows-style paths <driveletter>: Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: serial: fix GCC warningCorinna Vinschen2020-03-261-1/+1
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* newlib/libm/math: Make pow/powf return qnan for snan argKeith Packard via Newlib2020-03-262-7/+16
| | | | | | | | The IEEE spec for pow only has special case for x**0 and 1**y when x/y are quiet NaN. For signaling NaN, the general case applies and these functions should signal the invalid exception and return a quiet NaN. Signed-off-by: Keith Packard <keithp@keithp.com>
* newlib/libm/common: Don't re-convert float to bits in modf/modffKeith Packard via Newlib2020-03-262-15/+5
| | | | | | | | These functions shared a pattern of re-converting the argument to bits when returning +/-0. Skip that as the initial conversion still has the sign bit. Signed-off-by: Keith Packard <keithp@keithp.com>
* newlib/libm/common: Fix modf/modff returning snanKeith Packard via Newlib2020-03-262-16/+4
| | | | | | | | | Recent GCC appears to elide multiplication by 1, which causes snan parameters to be returned unchanged through *iptr. Use the existing conversion of snan to qnan to also set the correct result in *iptr instead. Signed-off-by: Keith Packard <keithp@keithp.com>
* Fix spurious underflow exceptions for Bessel functions for double(from glibc ↵Joseph S. Myers2020-03-264-9/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug 14155) This fix comes from glibc, from files which originated from the same place as the newlib files. Those files in glibc carry the same license as the newlib files. Bug 14155 is spurious underflow exceptions from Bessel functions for large arguments. (The correct results for large x are roughly constant * sin or cos (x + constant) / sqrt (x), so no underflow exceptions should occur based on the final result.) There are various places underflows may occur in the intermediate calculations that cause the failures listed in that bug. This patch fixes problems for the double version where underflows occur in calculating the intermediate functions P and Q (in particular, x**-12 gets computed while calculating Q). Appropriate approximations are used for P and Q for arguments at least 0x1p28 and above to avoid the underflows. For sufficiently large x - 0x1p129 and above - the code already has a cut-off to avoid calculating P and Q at all, which means the approximations -0.125 / x and 0.375 / x can't themselves cause underflows calculating Q. This cut-off is heuristically reasonable for the point beyond which Q can be neglected (based on expecting around 0x1p-64 to be the least absolute value of sin or cos for large arguments representable in double). The float versions use a cut-off 0x1p17, which is less heuristically justifiable but should still only affect values near zeroes of the Bessel functions where these implementations are intrinsically inaccurate anyway (bugs 14469-14472), and should serve to avoid underflows (the float underflow for jn in bug 14155 probably comes from the recurrence to compute jn). ldbl-96 uses 0x1p129, which may not really be enough heuristically (0x1p143 or so might be safer - 143 = 64 + 79, number of mantissa bits plus total number of significant bits in representation) but again should avoid underflows and only affect values where the code is substantially inaccurate anyway. ldbl-128 and ldbl-128ibm share a completely different implementation with no such cut-off, which I propose to fix separately. Signed-off-by: Keith Packard <keithp@keithp.com>
* Revert "Cygwin: serial: read: if VMIN > 0, wait for VMIN chars in inbound queue"Corinna Vinschen2020-03-251-26/+12
| | | | | | | | | This reverts commit 082f2513c721e942d0fd563c4dc9117eee3513ab. Turns out, Linux as well as BSD really only wait for the smaller number, MIN or # of requested bytes. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* fhandler_serial: fix commentsCorinna Vinschen2020-03-251-3/+3
| | | | Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: serial: read: if VMIN > 0, wait for VMIN chars in inbound queueCorinna Vinschen2020-03-251-12/+26
| | | | | | | | Per termios, read waits for MIN chars even if the number of requested bytes is less. This requires to add WaitCommEvent to wait non-busily for MIN chars prior to calling ReadFile, so, reintroduce it. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
* Cygwin: serial: use per call OVERLAPPED structsCorinna Vinschen2020-03-234-81/+56
| | | | | | | | | | Sharing the OVERLAPPED struct and event object in there between read and select calls in the fhandler might have been a nice optimization way back when, but it is a dangerous, not thread-safe approach. Fix this by creating per-fhandler, per-call OVERLAPPED structs and event objects on demand. Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
generated by cgit v1.2.3 (git 2.25.1) at 2025-07-20 12:14:13 +0000