When creating volatile environment registry keys during first
wineprefix creation, none of the user folders are existent. Create
them before setting registry keys so that volatile environment gets
initialized and corresponding environment variables such as
%USERPROFILE% can be set. Otherwise, such environment variables are
set only after another wineboot, causing applications to fail if they
need them after first boot.
Signed-off-by: Zhiyi Zhang <zzhang@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When processing a (..) multiline section, each line is processed and
if it starts with a '@' it is not echoed, but more importantly if is
'rem' then anything else on that line should be ignored. The reported
issue was that a pipe was being executed when it was hidden behind a
rem, which was trigged by the preceeding '@' character not being
skipped.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45729
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
With the 'for' loop /f syntax, if tokens are requested the the normal
syntax is something like tokens=1,2* but there is valid syntax like
1,2,* (which effectively means the same). Make this other syntax work.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45722
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Avoid whitespace affecting the parsing of a for loops items. The
leading and trailing quote or backtick needed removing, and it was
assumed that the trailing character would be that character, which was
wrong when there was whitespace unless the parameter is trimmed.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45731
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When a batch label is called, %0 and %~0 should be the label being
called, and if you start adding modifiers to it (eg %~d0) then you get
details of the batch program containing the label.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=44369
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
'if exists' takes a parameter which can be directory, directory\ or
directory\. for example, and should equate to true if the directory
exists. The syntax directory\ is explicitly rejected by FindFirstFile
and hence was not working - look for this specific case, and if found
append a '.'.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45506
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
for /f can run a program and parse its output. The program name can
supply args and be quoted or not. If quoted, wine fails to run the
program because internally we were adding an extra pair of
quotes. These are not needed and can be removed.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=39906
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
A single line if statement causes problems when it has redirects
and/or continuation type operators (|, &&, || etc) because it is
expected that if there is more than one command in the 'if', then it
will use brackets. This patch changes the 'if' parsing to emulate
brackets at a continuation character. In addition, 'for' and 'if'
statements do not have their output redirected immediately, instead it
is redirected on the individual commands being executed not the
statement itself. We were opening the redirect once for the 'if' and
once for the processing of the statement inside the if.
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When 's' is used as a modifier, the paths that are presented to the
other modifiers needs to be a short path. Given the 'filename' part of
the path may not exist, we cannot use GetShortPathName directly
without first removing the filename part to just leave the directory
bit.
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Make that the registry entry
HKLM\Software\Microsoft\Windows NT\CurrentVersion\CurrentBuild
has always the same value as
HKLM\Software\Microsoft\Windows NT\CurrentVersion\CurrentBuildNumber
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=42870
Signed-off-by: Thomas Faller <tfaller1@gmx.de>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We can't implement ZwLoadDriver() on top of StartService(), since the latter
takes the service database lock. Instead simply move the entire body of
create_driver()/unload_driver() into ZwLoadDriver()/ZwUnloadDriver().
Similarly, clean up the list of loaded drivers in ntoskrnl rather than
winedevice.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45084
Signed-off-by: Zebediah Figura <z.figura12@gmail.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This essentially reverts 440482d2ef.
440482d was aimed towards making it possible to load multiple drivers
asynchronously, as well as to allow reentrancy. Unfortunately, asynchronicity
is incorrect, as demonstrated by bug 38836, and some trivial testing shows
that the SCM database lock is held for the entirety of the driver entry and
exit routines, and that StartService() and ControlService() block until they
complete. 5726824 and dd2624a nullified the effects of 440482d, making driver
loading all but synchronous (with the exception of the added 30 second
timeout, but this is actually incorrect: drivers can block indefinitely).
This patch therefore does not change any behaviour, but rather removes the
use of threadpools and "async" functions, essentially reverting back to the
implementation prior to 440482d. The incidental change to unload_driver()
made by that patch (viz. never to unload a driver without a DriverUnload()
routine) is kept.
Signed-off-by: Zebediah Figura <z.figura12@gmail.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
The Message TVM_EXPAND is captured in treeview_notify, then passes
g_pChildWnd->hTreeWnd on to OnTreeExpanding which hasn't been
initialized yet. The function get_last_key is already called after
the hTreeWnd has been assigned to select the previous selected
item (if it exists).
Signed-off-by: Alistair Leslie-Hughes <leslie_alistair@hotmail.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Since it could potentially change on us during an asynchronous custom action.
Signed-off-by: Zebediah Figura <z.figura12@gmail.com>
Signed-off-by: Hans Leidekker <hans@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Semicolons are also allowed inside a path, as long as they are quoted.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=45552
Signed-off-by: Fabian Maurer <dark.shadow4@web.de>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When inside a for loop, an 'if' statement is processed and the true part
taken. Once all the commands in the true are processed, the else part is
parsed, and a flag set to skip all commands in the else part. Unfortunately
this flag is left on even when the if statement ends, meaning subsequent
commands are also skipped.
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
When parsing a command, after the first '/' we store the characters away
in quals. The command itself can be MAXSTRING in bytes, but the quals was
limited to MAX_PATH. This is incorrect, as you can provide very long
qualifiers as well. Expand the space to allow the maximum size possible.
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
for /f allows a special syntax of tokens=* (rather than tokens=1* for example)
which just means put the whole line into the next variable). Note the handling of
the 'next variable' was wrong in the case of it being 'A' or 'a' as the wrap
calculation was wrong, but this only affected using this new syntax.
Signed-off-by: Jason Edmeades <us@edmeades.me.uk>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
We use a named pipe to communicate between the client (i.e. the process that
called MsiInstallProduct() and the custom action server. The naïve approach
has the client send custom action GUIDs to the server and the server send
back the results of executing the action, but this fails in the case of
nested custom actions, so instead we send back handles of threads for the
client to wait on.
Signed-off-by: Zebediah Figura <z.figura12@gmail.com>
Signed-off-by: Hans Leidekker <hans@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This patch is effectively a no-op by itself, but makes the next patch
architecturally much simpler. We need the main thread to be non-blocking
to allow nested custom actions to work, so creating the thread inside of
msiexec allows the custom action thread to write the result of the action to
the server pipe while the main thread simply reads from it.
Signed-off-by: Zebediah Figura <z.figura12@gmail.com>
Signed-off-by: Hans Leidekker <hans@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Alexandre Julliard