cosmopolitan/libc/calls/ntspawn.c

165 lines
6.8 KiB
C
Raw Normal View History

Make improvements - We now serialize the file descriptor table when spawning / executing processes on Windows. This means you can now inherit more stuff than just standard i/o. It's needed by bash, which duplicates the console to file descriptor #255. We also now do a better job serializing the environment variables, so you're less likely to encounter E2BIG when using your bash shell. We also no longer coerce environ to uppercase - execve() on Windows now remotely controls its parent process to make them spawn a replacement for itself. Then it'll be able to terminate immediately once the spawn succeeds, without having to linger around for the lifetime as a shell process for proxying the exit code. When process worker thread running in the parent sees the child die, it's given a handle to the new child, to replace it in the process table. - execve() and posix_spawn() on Windows will now provide CreateProcess an explicit handle list. This allows us to remove handle locks which enables better fork/spawn concurrency, with seriously correct thread safety. Other codebases like Go use the same technique. On the other hand fork() still favors the conventional WIN32 inheritence approach which can be a little bit messy, but is *controlled* by guaranteeing perfectly clean slates at both the spawning and execution boundaries - sigset_t is now 64 bits. Having it be 128 bits was a mistake because there's no reason to use that and it's only supported by FreeBSD. By using the system word size, signal mask manipulation on Windows goes very fast. Furthermore @asyncsignalsafe funcs have been rewritten on Windows to take advantage of signal masking, now that it's much more pleasant to use. - All the overlapped i/o code on Windows has been rewritten for pretty good signal and cancelation safety. We're now able to ensure overlap data structures are cleaned up so long as you don't longjmp() out of out of a signal handler that interrupted an i/o operation. Latencies are also improved thanks to the removal of lots of "busy wait" code. Waits should be optimal for everything except poll(), which shall be the last and final demon we slay in the win32 i/o horror show. - getrusage() on Windows is now able to report RUSAGE_CHILDREN as well as RUSAGE_SELF, thanks to aggregation in the process manager thread.
2023-10-08 12:36:18 +00:00
/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:2;tab-width:8;coding:utf-8 -*-│
vi: set net ft=c ts=2 sts=2 sw=2 fenc=utf-8 :vi
Copyright 2021 Justine Alexandra Roberts Tunney
Permission to use, copy, modify, and/or distribute this software for
any purpose with or without fee is hereby granted, provided that the
above copyright notice and this permission notice appear in all copies.
THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE
AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.
*/
#include "libc/proc/ntspawn.h"
#include "libc/calls/struct/sigset.internal.h"
#include "libc/calls/syscall_support-nt.internal.h"
#include "libc/intrin/strace.internal.h"
#include "libc/nt/enum/processaccess.h"
#include "libc/nt/enum/processcreationflags.h"
#include "libc/nt/errors.h"
#include "libc/nt/files.h"
#include "libc/nt/memory.h"
#include "libc/nt/process.h"
#include "libc/nt/runtime.h"
#include "libc/nt/startupinfo.h"
#include "libc/nt/struct/processinformation.h"
#include "libc/nt/struct/procthreadattributelist.h"
#include "libc/nt/struct/startupinfo.h"
#include "libc/nt/struct/startupinfoex.h"
#include "libc/proc/ntspawn.h"
#include "libc/str/str.h"
#include "libc/sysv/errfuns.h"
#ifdef __x86_64__
struct SpawnBlock {
char16_t path[PATH_MAX];
char16_t cmdline[32767];
char16_t envblock[32767];
char envbuf[32767];
};
static void *ntspawn_malloc(size_t size) {
return HeapAlloc(GetProcessHeap(), 0, size);
}
static void ntspawn_free(void *ptr) {
HeapFree(GetProcessHeap(), 0, ptr);
}
/**
* Spawns process on Windows NT.
*
* This function delegates to CreateProcess() with UTF-8 UTF-16
* translation and argv escaping. Please note this will NOT escape
* command interpreter syntax.
*
* @param prog won't be PATH searched
* @param argv specifies prog arguments
* @param envp[𝟶,m-2] specifies "foo=bar" environment variables, which
* don't need to be passed in sorted order; however, this function
* goes faster the closer they are to sorted
* @param envp[m-1] is NULL
* @param extravars is added to envp to avoid setenv() in caller
* @param opt_out_lpProcessInformation can be used to return process and
* thread IDs to parent, as well as open handles that need close()
* @return 0 on success, or -1 w/ errno
* @see spawnve() which abstracts this function
* @asyncsignalsafe
*/
textwindows int ntspawn(
int64_t dirhand, const char *prog, char *const argv[], char *const envp[],
Make improvements - We now serialize the file descriptor table when spawning / executing processes on Windows. This means you can now inherit more stuff than just standard i/o. It's needed by bash, which duplicates the console to file descriptor #255. We also now do a better job serializing the environment variables, so you're less likely to encounter E2BIG when using your bash shell. We also no longer coerce environ to uppercase - execve() on Windows now remotely controls its parent process to make them spawn a replacement for itself. Then it'll be able to terminate immediately once the spawn succeeds, without having to linger around for the lifetime as a shell process for proxying the exit code. When process worker thread running in the parent sees the child die, it's given a handle to the new child, to replace it in the process table. - execve() and posix_spawn() on Windows will now provide CreateProcess an explicit handle list. This allows us to remove handle locks which enables better fork/spawn concurrency, with seriously correct thread safety. Other codebases like Go use the same technique. On the other hand fork() still favors the conventional WIN32 inheritence approach which can be a little bit messy, but is *controlled* by guaranteeing perfectly clean slates at both the spawning and execution boundaries - sigset_t is now 64 bits. Having it be 128 bits was a mistake because there's no reason to use that and it's only supported by FreeBSD. By using the system word size, signal mask manipulation on Windows goes very fast. Furthermore @asyncsignalsafe funcs have been rewritten on Windows to take advantage of signal masking, now that it's much more pleasant to use. - All the overlapped i/o code on Windows has been rewritten for pretty good signal and cancelation safety. We're now able to ensure overlap data structures are cleaned up so long as you don't longjmp() out of out of a signal handler that interrupted an i/o operation. Latencies are also improved thanks to the removal of lots of "busy wait" code. Waits should be optimal for everything except poll(), which shall be the last and final demon we slay in the win32 i/o horror show. - getrusage() on Windows is now able to report RUSAGE_CHILDREN as well as RUSAGE_SELF, thanks to aggregation in the process manager thread.
2023-10-08 12:36:18 +00:00
char *const extravars[], uint32_t dwCreationFlags,
const char16_t *opt_lpCurrentDirectory, int64_t opt_hParentProcess,
int64_t *opt_lpExplicitHandleList, uint32_t dwExplicitHandleCount,
const struct NtStartupInfo *lpStartupInfo,
struct NtProcessInformation *opt_out_lpProcessInformation) {
int rc = -1;
struct SpawnBlock *sb;
BLOCK_SIGNALS;
if ((sb = ntspawn_malloc(sizeof(*sb))) &&
__mkntpathath(dirhand, prog, 0, sb->path) != -1) {
Make improvements - We now serialize the file descriptor table when spawning / executing processes on Windows. This means you can now inherit more stuff than just standard i/o. It's needed by bash, which duplicates the console to file descriptor #255. We also now do a better job serializing the environment variables, so you're less likely to encounter E2BIG when using your bash shell. We also no longer coerce environ to uppercase - execve() on Windows now remotely controls its parent process to make them spawn a replacement for itself. Then it'll be able to terminate immediately once the spawn succeeds, without having to linger around for the lifetime as a shell process for proxying the exit code. When process worker thread running in the parent sees the child die, it's given a handle to the new child, to replace it in the process table. - execve() and posix_spawn() on Windows will now provide CreateProcess an explicit handle list. This allows us to remove handle locks which enables better fork/spawn concurrency, with seriously correct thread safety. Other codebases like Go use the same technique. On the other hand fork() still favors the conventional WIN32 inheritence approach which can be a little bit messy, but is *controlled* by guaranteeing perfectly clean slates at both the spawning and execution boundaries - sigset_t is now 64 bits. Having it be 128 bits was a mistake because there's no reason to use that and it's only supported by FreeBSD. By using the system word size, signal mask manipulation on Windows goes very fast. Furthermore @asyncsignalsafe funcs have been rewritten on Windows to take advantage of signal masking, now that it's much more pleasant to use. - All the overlapped i/o code on Windows has been rewritten for pretty good signal and cancelation safety. We're now able to ensure overlap data structures are cleaned up so long as you don't longjmp() out of out of a signal handler that interrupted an i/o operation. Latencies are also improved thanks to the removal of lots of "busy wait" code. Waits should be optimal for everything except poll(), which shall be the last and final demon we slay in the win32 i/o horror show. - getrusage() on Windows is now able to report RUSAGE_CHILDREN as well as RUSAGE_SELF, thanks to aggregation in the process manager thread.
2023-10-08 12:36:18 +00:00
if (!mkntcmdline(sb->cmdline, argv) &&
!mkntenvblock(sb->envblock, envp, extravars, sb->envbuf)) {
bool32 ok;
int64_t dp = GetCurrentProcess();
// create attribute list
// this code won't call malloc in practice
void *freeme = 0;
_Alignas(16) char memory[128];
size_t size = sizeof(memory);
struct NtProcThreadAttributeList *alist = (void *)memory;
uint32_t items = !!opt_hParentProcess + !!opt_lpExplicitHandleList;
ok = InitializeProcThreadAttributeList(alist, items, 0, &size);
if (!ok && GetLastError() == kNtErrorInsufficientBuffer) {
ok = !!(alist = freeme = ntspawn_malloc(size));
if (ok) {
ok = InitializeProcThreadAttributeList(alist, items, 0, &size);
}
}
if (ok && opt_hParentProcess) {
ok = UpdateProcThreadAttribute(
alist, 0, kNtProcThreadAttributeParentProcess, &opt_hParentProcess,
sizeof(opt_hParentProcess), 0, 0);
}
if (ok && opt_lpExplicitHandleList) {
ok = UpdateProcThreadAttribute(
alist, 0, kNtProcThreadAttributeHandleList,
opt_lpExplicitHandleList,
dwExplicitHandleCount * sizeof(*opt_lpExplicitHandleList), 0, 0);
}
// create the process
if (ok) {
struct NtStartupInfoEx info;
bzero(&info, sizeof(info));
info.StartupInfo = *lpStartupInfo;
info.StartupInfo.cb = sizeof(info);
info.lpAttributeList = alist;
if (ok) {
if (CreateProcess(sb->path, sb->cmdline, 0, 0, true,
dwCreationFlags | kNtCreateUnicodeEnvironment |
kNtExtendedStartupinfoPresent |
kNtInheritParentAffinity,
sb->envblock, opt_lpCurrentDirectory,
&info.StartupInfo, opt_out_lpProcessInformation)) {
rc = 0;
} else {
STRACE("CreateProcess() failed w/ %d", GetLastError());
if (GetLastError() == kNtErrorSharingViolation) {
etxtbsy();
} else if (GetLastError() == kNtErrorInvalidName) {
enoent();
Make improvements - We now serialize the file descriptor table when spawning / executing processes on Windows. This means you can now inherit more stuff than just standard i/o. It's needed by bash, which duplicates the console to file descriptor #255. We also now do a better job serializing the environment variables, so you're less likely to encounter E2BIG when using your bash shell. We also no longer coerce environ to uppercase - execve() on Windows now remotely controls its parent process to make them spawn a replacement for itself. Then it'll be able to terminate immediately once the spawn succeeds, without having to linger around for the lifetime as a shell process for proxying the exit code. When process worker thread running in the parent sees the child die, it's given a handle to the new child, to replace it in the process table. - execve() and posix_spawn() on Windows will now provide CreateProcess an explicit handle list. This allows us to remove handle locks which enables better fork/spawn concurrency, with seriously correct thread safety. Other codebases like Go use the same technique. On the other hand fork() still favors the conventional WIN32 inheritence approach which can be a little bit messy, but is *controlled* by guaranteeing perfectly clean slates at both the spawning and execution boundaries - sigset_t is now 64 bits. Having it be 128 bits was a mistake because there's no reason to use that and it's only supported by FreeBSD. By using the system word size, signal mask manipulation on Windows goes very fast. Furthermore @asyncsignalsafe funcs have been rewritten on Windows to take advantage of signal masking, now that it's much more pleasant to use. - All the overlapped i/o code on Windows has been rewritten for pretty good signal and cancelation safety. We're now able to ensure overlap data structures are cleaned up so long as you don't longjmp() out of out of a signal handler that interrupted an i/o operation. Latencies are also improved thanks to the removal of lots of "busy wait" code. Waits should be optimal for everything except poll(), which shall be the last and final demon we slay in the win32 i/o horror show. - getrusage() on Windows is now able to report RUSAGE_CHILDREN as well as RUSAGE_SELF, thanks to aggregation in the process manager thread.
2023-10-08 12:36:18 +00:00
}
}
rc = __fix_enotdir(rc, sb->path);
}
} else {
rc = __winerr();
}
// clean up resources
if (alist) {
DeleteProcThreadAttributeList(alist);
}
if (freeme) {
ntspawn_free(freeme);
}
if (dp && dp != GetCurrentProcess()) {
CloseHandle(dp);
}
}
}
if (sb) ntspawn_free(sb);
ALLOW_SIGNALS;
return rc;
}
#endif /* __x86_64__ */