Since VS 16.11 the compiler sometimes emits calls to _RTC_UninitUse, when parts of a bitfield are initialized (See https://developercommunity.visualstudio.com/t/Broken-runtime-checks-with-CL-19293013/1503629). Fix this by using an ULONG instead of a bitfield.
Note: The structure uses a 24 bit bitfield plus an UCHAR, which is supposed to form a 32 bit field, but that doesn't work anyway.
- usetup: New bootsector page.
- shell32: Copy and paste, and moving elements.
Also, some strings related to the shutdown and logoff.
- Minor Spanish grammar fix - some female words and minor latin american typos.
- First revision of the .inf, that includes the translation of the Services,
audio, processors and other drivers and minor things.
These drivers are based on NT4 DDK sample code, were originally
started by Alexey Bragin and then constantly patched by Pierre
Schweitzer to fix compatibility with NT5+ storage stack.
Replaced with Microsoft drivers published on GitHub by an open
license.
These drivers were originally added as part of 4e7b22b216
The upstream driver is not maintained and the file system itself
is in a semi-abandoned state.
Originally imported at 3a3ef631d1
The driver is written by Lee Jae-Hong, updated by Bo Brantén.
ReactOS porting made by Peter Hater and Pierre Schweitzer.
Follow updates at http://www.acc.umu.se/~bosse/
FS Recognizer code is left to keep the FS support as an
installable driver.
CORE-11040
The upstream driver is not maintained and the file system itself
is in a semi-abandoned state.
Originally imported at e308102f4a
The driver is written by Mark W Piper, updated by Bo Brantén.
ReactOS porting made by Peter Hater and Pierre Schweitzer.
Follow updates at http://www.acc.umu.se/~bosse/
FS Recognizer code is left to keep the FS support as an
installable driver.
CORE-11005
This turns off mountmgr listening on new devices. Now
BTRFS will only be mounted on demand when OS asks for it. RAID
is not going to work this way probably (have anybody checked it at all?)
This is a temporary workaround for CORE-17469
Spinlocks are not reentrant (and this is done a lot), using them forces us to have
an horrible hack in the kernel, which unschedules threads which are at DISPATCH_LEVEL
thus allowing another thread to take ownership of the spinlock while the unscheduled
thread should already hold it.
CORE-6473
This speeds up boot a bit and also decreases binary size.
- Introduce a new CMake define ISAPNP_ENABLE that defined only in x86 platform
when ISA PnP is supported.
Rework the hardware support code to avoid unneeded transitions
between configuration states.
- Only call WaitForKey when needed.
- Each PnP state transition causes the bus to move to WaitForKey state.
Besides, deactivate logical device during processing remove request.
- Fix empty resource descriptors being created.
- Properly support IRQ descriptors.
- Introduce four helpers made to help search descriptors
in the logical device's requirements.
- Implement support for memory descriptors and alternative configurations.
- DMA descriptors are always DMA_8.
- Support all resource descriptors.
- Optimize card identification.
- Detect cards that is no longer present on the bus.
- Deactivate cards after the identification phase; they will be activated
by start device IRP.
- Provide a device description and compatible IDs to the device manager.
- Prevent duplicate IDs across multiple logical devices.
- Suppress warning about the usage of literals in port addresses.
- Implement device removal.
- Finish the Read Port special handling
introduced in 21514e473f.
- Gracefully handle low memory situations.
- Implement IRQ translator.
- Correctly stub power IRP handling.
- Remove unneeded dispatch routines; implement WMI IRP handling.
- Do not handle requests like WRITE_CONFIG.
- Add a signature member to device extensions to make checks clearer.
This PDO is created only once during start of first FDO.
Other buses will remain in an inactive state until
the first FDO receives a remove request.
CORE-17034
- Add SAL2 annotations.
- Use RTL_NUMBER_OF macro in kernel mode.
- Use better function/member names.
- Less hardcoded values.
- Add license information. See commit
d9face83c6 for details.
No functional changes.
I am not sure why this line was added.
I don't know how I could improve that case.
Then, disable it, for the time being.
Addendum to 4b9cf2e (r71252). CORE-12441 CORE-17371
Some virtual machines errorneously expose null PCI device function
on PIIX4 chipset where an USB controller should normally reside.
Windows pci.sys driver does not enumerate these devices.
Affected virtual machines:
- Connectix Virtual PC 5.1.370
- Connectix Virtual PC 5.2.418
- Microsoft Virtual PC 2004 SP1 (5.3.582.27)
- Microsoft Virtual PC 2007 SP1 (6.0.156.0)
Non-affected virtual machines:
- Connectix Virtual PC 4.0
- Microsoft Hyper-V
Fixes CORE-17636.
Do not treat target device change notification as
DEVICE_INTERFACE_CHANGE_NOTIFICATION. The notification have to be
unregistered while handling GUID_DEVICE_INTERFACE_REMOVAL, so
GUID_TARGET_DEVICE_REMOVE_COMPLETE should never be sent to mountmgr in a
normal case.
CORE-16106
- Do not access uninitialized SymlinkInformation on failure case
and just return
- Do not make an assumption that MOUNTMGR_TARGET_NAME has a zero-string
From MSDN:
It is an error to call KeReleaseSpinLockFromDpcLevel if the specified spin lock was acquired by calling KeAcquireSpinLock because the caller's original IRQL is not restored, which can cause deadlocks or fatal page faults.
- Use PnP storage class drivers
- Make partmgr an upper filter driver for Disk class
- Fill upper filters in txtsetup and usetup/devinst
- Add cdrom driver to the critical device database
CORE-6264
- Pass all SRB flags which Windows scsiport passes
- Correctly reset the queue after completion
This fixes the bug when MS cdrom driver hangs after media ejection
Basic functions are implemented in order to work in PnP stack,
only legacy (non-pnp) miniport drivers are supported.
Tested mostly with uniata
CORE-17132
This driver works as complement to disk.sys/classpnp.sys from Windows 10
Manages partition PDOs and exposes them as volumes to mountmgr.sys.
The driver is almost complete, just some minor IOCTLs missing (will be
added on demand)
- ReparseFile was concatenated with itself, instead of ReparseIndex
- Meanwhile, use RtlAppendUnicodeStringToString for concatenating
strings instead of raw memory operations
This function may stuck during device installation if there are issues
with interrupts (or with a device itself).
This fixes the boot on my testing ThinkPad x60s
- Change INIT_FUNCTION and INIT_SECTION to CODE_SEG("INIT") and DATA_SEG("INIT") respectively
- Remove INIT_FUNCTION from function prototypes
- Remove alloc_text pragma calls as they are not needed anymore
This avoids blocking all Ex worker threads in fastfat, thereby making Cc
unable to issue the lazy writes that would unblock those workers.
This is more or less directly taken from fastfat_new.
in favor of add_compile_options and the like with generator expressions
Also take this as an opportunity to remove the C++11 standard hack, GCC 8 now defaults to C++14
Instead of messing with global variables and the like, we introduce two target properties:
- WITH_CXX_EXCEPTIONS: if you want to use C++ exceptions
- WITH_CXX_RTTI: if you need RTTI in your module
You can use the newly introduced set_target_cpp_properties function, with WITH_EXCEPTIONS and WITH_RTTI arguments
We also introduce two libraries :
- cpprt: for C++ runtime routines
- cppstl: for the C++ standard template library
NB: On GCC, this requires to create imported libraries with the related built-in libraries:libsupc++, limingwex, libstdc++
Finally, we manage the relevant flags with the ad-hoc generator expressions
So, if you don't need exceptions, nor RTTI, nor use any runtime at all: you simply have nothing else to do than add your C++ file to your module
Cc may decide to process deferred writes any time, so the context might
already be freed by the time we return from CcDeferWrite.
Also mark the IRP as pending, since we're going to return STATUS_PENDING.
The source code is licensed under MS-PL license, taken from Windows Driver Samples
repository (microsoft/Windows-driver-samples@master/storage/class/cdrom/)
Synched with commit 96eb96dfb613e4c745db6bd1f53a92fe7e2290fc
The driver is written for Windows 10 and uses KMDF so we compile it with ntoskrnl_vista
and wdf01000 statically linked (for wdf01000 this will likely be changed in future)
CORE-17129
It adds basic input support for:
- Standard Bus Mouse
- Standard InPort Mouse
- Logitech Bus Mouse
- Microsoft Bus Mouse
- Microsoft InPort Mouse
- NEC PC-98 Bus Mouse
Untested on PC/AT, but should work.
Replace call to AllocatedBufferSize(), with BufferSize().
Indeed (quoting https://docs.microsoft.com/en-us/windows-hardware/drivers/ddi/portcls/nf-portcls-idmachannel-buffersize ):
> The BufferSize() method returns the buffer size that was set by the previous call to IDmaChannel::SetBufferSize(). If SetBufferSize() has not been called since the IDmaChannel::AllocateBuffer() call, BufferSize returns the allocated buffer size. The DMA-channel object does not actually use this value internally. This value is maintained by the object to **allow its various clients to communicate the intended size of the buffer**.
And this is exactly what we want to do.
The source code is licensed under MS-PL license, taken from Windows Driver Samples
repository (https://github.com/microsoft/Windows-driver-samples/tree/master/storage/class/disk/)
Synched with commit 3428c5feaac7c1a5e2a09db0ed93392f7568f9e6
The driver is written for Windows 8+, so we compile it with ntoskrnl_vista
statically linked and with NTDDI_WIN8 defined
CORE-17129
The source code is licensed under MS-PL license, taken from Windows Driver Samples
repository (https://github.com/microsoft/Windows-driver-samples/tree/master/storage/class/classpnp/)
Synched with commit 88541f70c4273ecd30c8c7c72135bc038a00fd88
The driver is written for Windows 8+, so we compile it with ntoskrnl_vista
statically linked and with NTDDI_WIN8 defined
CORE-17129
'...\write.c(38): error C2373: 'RfsdFloppyFlush': redefinition; different type modifiers'
'...\write.c(45): error C2373: 'RfsdFloppyFlushDpc': redefinition; different type modifiers'
It started to spam when more components of the MountMgr
were coded during 0.4.14dev.
According to Victor Perevertkin it is not crucial for us
to see it, as those are 'optional MountMgr features'.
Imho this points towards unimplemented stuff.
No official ros release has been affected, because I did
revert most of the new MountMgr features for 0.4.14release
earlier.
Some places have been changed to return FILE_DOES_NOT_EXIST, everything
else just returns 0. FastFAT driver from MS does the same (tm),
so I'm not going to have a big research here.
CORE-16623
Co-authored-by: Suraj K Suresh <kssuraj15@gmail.com>
We must not do anything on PDO removal, especially not deleting it.
We are allowed to delete it, only once we successfully completed a IRP_MN_QUERY_DEVICE_RELATIONS without with PDO.
JIRA-user "Illen" reported booting from his Z170 controller worked up to
0.4.12-dev-936-g89aaf0e
and would refuse booting - beginning with uniata commit
0.4.12-dev-937-g
b546130731
For sure this workaround is just a temporary and no proper solution,
but was confirmed to be working by "Illen".
We have no clear understanding of the real bug yet.
Can be replaced by something better at any time.
It was already committed into 0.4.12, 0.4.13, 0.4.14.
We never had an affected release therefore.
Since no one took care of this bug ever,
the workaround will now be committed to master as well.
cherry picked from commit 0.4.13-RC-9-g
11178f38e4
- Add boot video color constants
- Refactor palette initialization
- Move some common stuff in right place
- Get rid of some magic constants and hardcoded values
- Get rid of TopDelta variable (calculated at compile time)
- Update SAL annotations
Addendum to 5f2ca473. CORE-16216 CORE-16219
According to PC-9801 Bible p. 50, divisor for PIT will become unsupported in some cases after having removed the fractional part. Replace 19200 value with 9600 which is supported by both 10 MHz and 8 MHz machines.
When gdb receives a fault, it converts it to "signal", and send "C"
command to server to transfer the signal it couldn't handle.
On ReactOS (and Windows ?) side, we tell KD API that we continue with an
error code, so the exception handler can be called.
This is useful when playing with gflags, especially +soe, with KDGDB.
In ACPI resource descriptors, alternatives are marked with
StartDependent tags. Only the last set is terminated with EndDependent.
Therefore, since we only return the first alternative list for now,
ignore the first StartDependent tag and terminate enumeration at the second.
In the future we will need to build the full set of alternative lists here,
which will also make the unit test succeed fully.
This should fix random resource conflicts and make COM ports usable.
