mirror of
https://git.tartarus.org/simon/putty.git
synced 2025-03-20 05:45:27 -05:00

This brings various concrete advantages over the previous system: - consistent support for out-of-tree builds on all platforms - more thorough support for Visual Studio IDE project files - support for Ninja-based builds, which is particularly useful on Windows where the alternative nmake has no parallel option - a really simple set of build instructions that work the same way on all the major platforms (look how much shorter README is!) - better decoupling of the project configuration from the toolchain configuration, so that my Windows cross-building doesn't need (much) special treatment in CMakeLists.txt - configure-time tests on Windows as well as Linux, so that a lot of ad-hoc #ifdefs second-guessing a particular feature's presence from the compiler version can now be replaced by tests of the feature itself Also some longer-term software-engineering advantages: - other people have actually heard of CMake, so they'll be able to produce patches to the new build setup more easily - unlike the old mkfiles.pl, CMake is not my personal problem to maintain - most importantly, mkfiles.pl was just a horrible pile of unmaintainable cruft, which even I found it painful to make changes to or to use, and desperately needed throwing in the bin. I've already thrown away all the variants of it I had in other projects of mine, and was only delaying this one so we could make the 0.75 release branch first. This change comes with a noticeable build-level restructuring. The previous Recipe worked by compiling every object file exactly once, and then making each executable by linking a precisely specified subset of the same object files. But in CMake, that's not the natural way to work - if you write the obvious command that puts the same source file into two executable targets, CMake generates a makefile that compiles it once per target. That can be an advantage, because it gives you the freedom to compile it differently in each case (e.g. with a #define telling it which program it's part of). But in a project that has many executable targets and had carefully contrived to _never_ need to build any module more than once, all it does is bloat the build time pointlessly! To avoid slowing down the build by a large factor, I've put most of the modules of the code base into a collection of static libraries organised vaguely thematically (SSH, other backends, crypto, network, ...). That means all those modules can still be compiled just once each, because once each library is built it's reused unchanged for all the executable targets. One upside of this library-based structure is that now I don't have to manually specify exactly which objects go into which programs any more - it's enough to specify which libraries are needed, and the linker will figure out the fine detail automatically. So there's less maintenance to do in CMakeLists.txt when the source code changes. But that reorganisation also adds fragility, because of the trad Unix linker semantics of walking along the library list once each, so that cyclic references between your libraries will provoke link errors. The current setup builds successfully, but I suspect it only just manages it. (In particular, I've found that MinGW is the most finicky on this score of the Windows compilers I've tried building with. So I've included a MinGW test build in the new-look Buildscr, because otherwise I think there'd be a significant risk of introducing MinGW-only build failures due to library search order, which wasn't a risk in the previous library-free build organisation.) In the longer term I hope to be able to reduce the risk of that, via gradual reorganisation (in particular, breaking up too-monolithic modules, to reduce the risk of knock-on references when you included a module for function A and it also contains function B with an unsatisfied dependency you didn't really need). Ideally I want to reach a state in which the libraries all have sensibly described purposes, a clearly documented (partial) order in which they're permitted to depend on each other, and a specification of what stubs you have to put where if you're leaving one of them out (e.g. nocrypto) and what callbacks you have to define in your non-library objects to satisfy dependencies from things low in the stack (e.g. out_of_memory()). One thing that's gone completely missing in this migration, unfortunately, is the unfinished MacOS port linked against Quartz GTK. That's because it turned out that I can't currently build it myself, on my own Mac: my previous installation of GTK had bit-rotted as a side effect of an Xcode upgrade, and I haven't yet been able to persuade jhbuild to make me a new one. So I can't even build the MacOS port with the _old_ makefiles, and hence, I have no way of checking that the new ones also work. I hope to bring that port back to life at some point, but I don't want it to block the rest of this change.
468 lines
13 KiB
C
468 lines
13 KiB
C
/*
|
|
* winmisc.c: miscellaneous Windows-specific things
|
|
*/
|
|
|
|
#include <stdio.h>
|
|
#include <stdlib.h>
|
|
#include <limits.h>
|
|
#include "putty.h"
|
|
#ifndef SECURITY_WIN32
|
|
#define SECURITY_WIN32
|
|
#endif
|
|
#include <security.h>
|
|
|
|
DWORD osMajorVersion, osMinorVersion, osPlatformId;
|
|
|
|
char *platform_get_x_display(void) {
|
|
/* We may as well check for DISPLAY in case it's useful. */
|
|
return dupstr(getenv("DISPLAY"));
|
|
}
|
|
|
|
Filename *filename_from_str(const char *str)
|
|
{
|
|
Filename *ret = snew(Filename);
|
|
ret->path = dupstr(str);
|
|
return ret;
|
|
}
|
|
|
|
Filename *filename_copy(const Filename *fn)
|
|
{
|
|
return filename_from_str(fn->path);
|
|
}
|
|
|
|
const char *filename_to_str(const Filename *fn)
|
|
{
|
|
return fn->path;
|
|
}
|
|
|
|
bool filename_equal(const Filename *f1, const Filename *f2)
|
|
{
|
|
return !strcmp(f1->path, f2->path);
|
|
}
|
|
|
|
bool filename_is_null(const Filename *fn)
|
|
{
|
|
return !*fn->path;
|
|
}
|
|
|
|
void filename_free(Filename *fn)
|
|
{
|
|
sfree(fn->path);
|
|
sfree(fn);
|
|
}
|
|
|
|
void filename_serialise(BinarySink *bs, const Filename *f)
|
|
{
|
|
put_asciz(bs, f->path);
|
|
}
|
|
Filename *filename_deserialise(BinarySource *src)
|
|
{
|
|
return filename_from_str(get_asciz(src));
|
|
}
|
|
|
|
char filename_char_sanitise(char c)
|
|
{
|
|
if (strchr("<>:\"/\\|?*", c))
|
|
return '.';
|
|
return c;
|
|
}
|
|
|
|
char *get_username(void)
|
|
{
|
|
DWORD namelen;
|
|
char *user;
|
|
bool got_username = false;
|
|
DECL_WINDOWS_FUNCTION(static, BOOLEAN, GetUserNameExA,
|
|
(EXTENDED_NAME_FORMAT, LPSTR, PULONG));
|
|
|
|
{
|
|
static bool tried_usernameex = false;
|
|
if (!tried_usernameex) {
|
|
/* Not available on Win9x, so load dynamically */
|
|
HMODULE secur32 = load_system32_dll("secur32.dll");
|
|
/* If MIT Kerberos is installed, the following call to
|
|
GET_WINDOWS_FUNCTION makes Windows implicitly load
|
|
sspicli.dll WITHOUT proper path sanitizing, so better
|
|
load it properly before */
|
|
HMODULE sspicli = load_system32_dll("sspicli.dll");
|
|
(void)sspicli; /* squash compiler warning about unused variable */
|
|
GET_WINDOWS_FUNCTION(secur32, GetUserNameExA);
|
|
tried_usernameex = true;
|
|
}
|
|
}
|
|
|
|
if (p_GetUserNameExA) {
|
|
/*
|
|
* If available, use the principal -- this avoids the problem
|
|
* that the local username is case-insensitive but Kerberos
|
|
* usernames are case-sensitive.
|
|
*/
|
|
|
|
/* Get the length */
|
|
namelen = 0;
|
|
(void) p_GetUserNameExA(NameUserPrincipal, NULL, &namelen);
|
|
|
|
user = snewn(namelen, char);
|
|
got_username = p_GetUserNameExA(NameUserPrincipal, user, &namelen);
|
|
if (got_username) {
|
|
char *p = strchr(user, '@');
|
|
if (p) *p = 0;
|
|
} else {
|
|
sfree(user);
|
|
}
|
|
}
|
|
|
|
if (!got_username) {
|
|
/* Fall back to local user name */
|
|
namelen = 0;
|
|
if (!GetUserName(NULL, &namelen)) {
|
|
/*
|
|
* Apparently this doesn't work at least on Windows XP SP2.
|
|
* Thus assume a maximum of 256. It will fail again if it
|
|
* doesn't fit.
|
|
*/
|
|
namelen = 256;
|
|
}
|
|
|
|
user = snewn(namelen, char);
|
|
got_username = GetUserName(user, &namelen);
|
|
if (!got_username) {
|
|
sfree(user);
|
|
}
|
|
}
|
|
|
|
return got_username ? user : NULL;
|
|
}
|
|
|
|
void dll_hijacking_protection(void)
|
|
{
|
|
/*
|
|
* If the OS provides it, call SetDefaultDllDirectories() to
|
|
* prevent DLLs from being loaded from the directory containing
|
|
* our own binary, and instead only load from system32.
|
|
*
|
|
* This is a protection against hijacking attacks, if someone runs
|
|
* PuTTY directly from their web browser's download directory
|
|
* having previously been enticed into clicking on an unwise link
|
|
* that downloaded a malicious DLL to the same directory under one
|
|
* of various magic names that seem to be things that standard
|
|
* Windows DLLs delegate to.
|
|
*
|
|
* It shouldn't break deliberate loading of user-provided DLLs
|
|
* such as GSSAPI providers, because those are specified by their
|
|
* full pathname by the user-provided configuration.
|
|
*/
|
|
static HMODULE kernel32_module;
|
|
DECL_WINDOWS_FUNCTION(static, BOOL, SetDefaultDllDirectories, (DWORD));
|
|
|
|
if (!kernel32_module) {
|
|
kernel32_module = load_system32_dll("kernel32.dll");
|
|
#if !HAVE_SETDEFAULTDLLDIRECTORIES
|
|
/* For older Visual Studio, this function isn't available in
|
|
* the header files to type-check */
|
|
GET_WINDOWS_FUNCTION_NO_TYPECHECK(
|
|
kernel32_module, SetDefaultDllDirectories);
|
|
#else
|
|
GET_WINDOWS_FUNCTION(kernel32_module, SetDefaultDllDirectories);
|
|
#endif
|
|
}
|
|
|
|
if (p_SetDefaultDllDirectories) {
|
|
/* LOAD_LIBRARY_SEARCH_SYSTEM32 and explicitly specified
|
|
* directories only */
|
|
p_SetDefaultDllDirectories(LOAD_LIBRARY_SEARCH_SYSTEM32 |
|
|
LOAD_LIBRARY_SEARCH_USER_DIRS);
|
|
}
|
|
}
|
|
|
|
void init_winver(void)
|
|
{
|
|
OSVERSIONINFO osVersion;
|
|
static HMODULE kernel32_module;
|
|
DECL_WINDOWS_FUNCTION(static, BOOL, GetVersionExA, (LPOSVERSIONINFO));
|
|
|
|
if (!kernel32_module) {
|
|
kernel32_module = load_system32_dll("kernel32.dll");
|
|
/* Deliberately don't type-check this function, because that
|
|
* would involve using its declaration in a header file which
|
|
* triggers a deprecation warning. I know it's deprecated (see
|
|
* below) and don't need telling. */
|
|
GET_WINDOWS_FUNCTION_NO_TYPECHECK(kernel32_module, GetVersionExA);
|
|
}
|
|
|
|
ZeroMemory(&osVersion, sizeof(osVersion));
|
|
osVersion.dwOSVersionInfoSize = sizeof (OSVERSIONINFO);
|
|
if (p_GetVersionExA && p_GetVersionExA(&osVersion)) {
|
|
osMajorVersion = osVersion.dwMajorVersion;
|
|
osMinorVersion = osVersion.dwMinorVersion;
|
|
osPlatformId = osVersion.dwPlatformId;
|
|
} else {
|
|
/*
|
|
* GetVersionEx is deprecated, so allow for it perhaps going
|
|
* away in future API versions. If it's not there, simply
|
|
* assume that's because Windows is too _new_, so fill in the
|
|
* variables we care about to a value that will always compare
|
|
* higher than any given test threshold.
|
|
*
|
|
* Normally we should be checking against the presence of a
|
|
* specific function if possible in any case.
|
|
*/
|
|
osMajorVersion = osMinorVersion = UINT_MAX; /* a very high number */
|
|
osPlatformId = VER_PLATFORM_WIN32_NT; /* not Win32s or Win95-like */
|
|
}
|
|
}
|
|
|
|
HMODULE load_system32_dll(const char *libname)
|
|
{
|
|
/*
|
|
* Wrapper function to load a DLL out of c:\windows\system32
|
|
* without going through the full DLL search path. (Hence no
|
|
* attack is possible by placing a substitute DLL earlier on that
|
|
* path.)
|
|
*/
|
|
static char *sysdir = NULL;
|
|
static size_t sysdirsize = 0;
|
|
char *fullpath;
|
|
HMODULE ret;
|
|
|
|
if (!sysdir) {
|
|
size_t len;
|
|
while ((len = GetSystemDirectory(sysdir, sysdirsize)) >= sysdirsize)
|
|
sgrowarray(sysdir, sysdirsize, len);
|
|
}
|
|
|
|
fullpath = dupcat(sysdir, "\\", libname);
|
|
ret = LoadLibrary(fullpath);
|
|
sfree(fullpath);
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
* A tree234 containing mappings from system error codes to strings.
|
|
*/
|
|
|
|
struct errstring {
|
|
int error;
|
|
char *text;
|
|
};
|
|
|
|
static int errstring_find(void *av, void *bv)
|
|
{
|
|
int *a = (int *)av;
|
|
struct errstring *b = (struct errstring *)bv;
|
|
if (*a < b->error)
|
|
return -1;
|
|
if (*a > b->error)
|
|
return +1;
|
|
return 0;
|
|
}
|
|
static int errstring_compare(void *av, void *bv)
|
|
{
|
|
struct errstring *a = (struct errstring *)av;
|
|
return errstring_find(&a->error, bv);
|
|
}
|
|
|
|
static tree234 *errstrings = NULL;
|
|
|
|
const char *win_strerror(int error)
|
|
{
|
|
struct errstring *es;
|
|
|
|
if (!errstrings)
|
|
errstrings = newtree234(errstring_compare);
|
|
|
|
es = find234(errstrings, &error, errstring_find);
|
|
|
|
if (!es) {
|
|
char msgtext[65536]; /* maximum size for FormatMessage is 64K */
|
|
|
|
es = snew(struct errstring);
|
|
es->error = error;
|
|
if (!FormatMessage((FORMAT_MESSAGE_FROM_SYSTEM |
|
|
FORMAT_MESSAGE_IGNORE_INSERTS), NULL, error,
|
|
MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
|
|
msgtext, lenof(msgtext)-1, NULL)) {
|
|
sprintf(msgtext,
|
|
"(unable to format: FormatMessage returned %u)",
|
|
(unsigned int)GetLastError());
|
|
} else {
|
|
int len = strlen(msgtext);
|
|
if (len > 0 && msgtext[len-1] == '\n')
|
|
msgtext[len-1] = '\0';
|
|
}
|
|
es->text = dupprintf("Error %d: %s", error, msgtext);
|
|
add234(errstrings, es);
|
|
}
|
|
|
|
return es->text;
|
|
}
|
|
|
|
FontSpec *fontspec_new(const char *name, bool bold, int height, int charset)
|
|
{
|
|
FontSpec *f = snew(FontSpec);
|
|
f->name = dupstr(name);
|
|
f->isbold = bold;
|
|
f->height = height;
|
|
f->charset = charset;
|
|
return f;
|
|
}
|
|
FontSpec *fontspec_copy(const FontSpec *f)
|
|
{
|
|
return fontspec_new(f->name, f->isbold, f->height, f->charset);
|
|
}
|
|
void fontspec_free(FontSpec *f)
|
|
{
|
|
sfree(f->name);
|
|
sfree(f);
|
|
}
|
|
void fontspec_serialise(BinarySink *bs, FontSpec *f)
|
|
{
|
|
put_asciz(bs, f->name);
|
|
put_uint32(bs, f->isbold);
|
|
put_uint32(bs, f->height);
|
|
put_uint32(bs, f->charset);
|
|
}
|
|
FontSpec *fontspec_deserialise(BinarySource *src)
|
|
{
|
|
const char *name = get_asciz(src);
|
|
unsigned isbold = get_uint32(src);
|
|
unsigned height = get_uint32(src);
|
|
unsigned charset = get_uint32(src);
|
|
return fontspec_new(name, isbold, height, charset);
|
|
}
|
|
|
|
bool open_for_write_would_lose_data(const Filename *fn)
|
|
{
|
|
WIN32_FILE_ATTRIBUTE_DATA attrs;
|
|
if (!GetFileAttributesEx(fn->path, GetFileExInfoStandard, &attrs)) {
|
|
/*
|
|
* Generally, if we don't identify a specific reason why we
|
|
* should return true from this function, we return false, and
|
|
* let the subsequent attempt to open the file for real give a
|
|
* more useful error message.
|
|
*/
|
|
return false;
|
|
}
|
|
if (attrs.dwFileAttributes & (FILE_ATTRIBUTE_DEVICE |
|
|
FILE_ATTRIBUTE_DIRECTORY)) {
|
|
/*
|
|
* File is something other than an ordinary disk file, so
|
|
* opening it for writing will not cause truncation. (It may
|
|
* not _succeed_ either, but that's not our problem here!)
|
|
*/
|
|
return false;
|
|
}
|
|
if (attrs.nFileSizeHigh == 0 && attrs.nFileSizeLow == 0) {
|
|
/*
|
|
* File is zero-length (or may be a named pipe, which
|
|
* dwFileAttributes can't tell apart from a regular file), so
|
|
* opening it for writing won't truncate any data away because
|
|
* there's nothing to truncate anyway.
|
|
*/
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
void escape_registry_key(const char *in, strbuf *out)
|
|
{
|
|
bool candot = false;
|
|
static const char hex[16] = "0123456789ABCDEF";
|
|
|
|
while (*in) {
|
|
if (*in == ' ' || *in == '\\' || *in == '*' || *in == '?' ||
|
|
*in == '%' || *in < ' ' || *in > '~' || (*in == '.'
|
|
&& !candot)) {
|
|
put_byte(out, '%');
|
|
put_byte(out, hex[((unsigned char) *in) >> 4]);
|
|
put_byte(out, hex[((unsigned char) *in) & 15]);
|
|
} else
|
|
put_byte(out, *in);
|
|
in++;
|
|
candot = true;
|
|
}
|
|
}
|
|
|
|
void unescape_registry_key(const char *in, strbuf *out)
|
|
{
|
|
while (*in) {
|
|
if (*in == '%' && in[1] && in[2]) {
|
|
int i, j;
|
|
|
|
i = in[1] - '0';
|
|
i -= (i > 9 ? 7 : 0);
|
|
j = in[2] - '0';
|
|
j -= (j > 9 ? 7 : 0);
|
|
|
|
put_byte(out, (i << 4) + j);
|
|
in += 3;
|
|
} else {
|
|
put_byte(out, *in++);
|
|
}
|
|
}
|
|
}
|
|
|
|
#ifdef DEBUG
|
|
static FILE *debug_fp = NULL;
|
|
static HANDLE debug_hdl = INVALID_HANDLE_VALUE;
|
|
static int debug_got_console = 0;
|
|
|
|
void dputs(const char *buf)
|
|
{
|
|
DWORD dw;
|
|
|
|
if (!debug_got_console) {
|
|
if (AllocConsole()) {
|
|
debug_got_console = 1;
|
|
debug_hdl = GetStdHandle(STD_OUTPUT_HANDLE);
|
|
}
|
|
}
|
|
if (!debug_fp) {
|
|
debug_fp = fopen("debug.log", "w");
|
|
}
|
|
|
|
if (debug_hdl != INVALID_HANDLE_VALUE) {
|
|
WriteFile(debug_hdl, buf, strlen(buf), &dw, NULL);
|
|
}
|
|
fputs(buf, debug_fp);
|
|
fflush(debug_fp);
|
|
}
|
|
#endif
|
|
|
|
char *registry_get_string(HKEY root, const char *path, const char *leaf)
|
|
{
|
|
HKEY key = root;
|
|
bool need_close_key = false;
|
|
char *toret = NULL, *str = NULL;
|
|
|
|
if (path) {
|
|
if (RegCreateKey(key, path, &key) != ERROR_SUCCESS)
|
|
goto out;
|
|
need_close_key = true;
|
|
}
|
|
|
|
DWORD type, size;
|
|
if (RegQueryValueEx(key, leaf, 0, &type, NULL, &size) != ERROR_SUCCESS)
|
|
goto out;
|
|
if (type != REG_SZ)
|
|
goto out;
|
|
|
|
str = snewn(size + 1, char);
|
|
DWORD size_got = size;
|
|
if (RegQueryValueEx(key, leaf, 0, &type, (LPBYTE)str,
|
|
&size_got) != ERROR_SUCCESS)
|
|
goto out;
|
|
if (type != REG_SZ || size_got > size)
|
|
goto out;
|
|
str[size_got] = '\0';
|
|
|
|
toret = str;
|
|
str = NULL;
|
|
|
|
out:
|
|
if (need_close_key)
|
|
RegCloseKey(key);
|
|
sfree(str);
|
|
return toret;
|
|
}
|