Equivalent of Linux sync_file_range system call on windows? - winapi

I need to fsync a range of byte I appended to a file without forcing a flush on the metadata (filesize,...).

As you've said it's ring3 and it's C++, here is the answer:
You need to call FlushViewOfFile after mapping the file. According to MSDN:
The FlushViewOfFile function does not flush the file metadata, and it
does not wait to return until the changes are flushed from the
underlying hardware disk cache and physically written to disk.
source: https://msdn.microsoft.com/en-us/library/windows/desktop/aa366563%28v=vs.85%29.aspx
An example code that writes data and uses FlushViewOfFile is here: http://forums.codeguru.com/showthread.php?367742-FlushViewOfFile-does-not-Flush

Related

Memory mapped file truncation on windows

Why can't we truncate memory mapped file on windows? Say I have mapped 4kb file and only used 2kb. So to truncate the 2kb, I need to first unmap the mapped file and then truncate it. why is this case? Is there a problem with API or my understanding is wrong?
Because on Windows once file region is mapped, it cannot be removed.
What happens when you map 32kb file into memory, then truncate it to 0. What should happen once you access that mapping? On POSIX systems memory access will generate a signal which will probably crash your process. On Windows this is simply not allowed.
This behavior is supposed to make programming with memory-mapped files easier as handling error code is easier that signal handling.
Because of this, CreateFileMapping extends file size to the size of created mapping object.

When Resources of a PE file are loaded

When using a resource included in a PE file (for example a binary resource) in C++ . we have to
first call
1 )FindResource and then
2 )LoadResource
to access the resource .
Being accurate about the function name "LoadResource" i wonder if the "Windows Loader" does load all resource of an application in memory just when loading other parts (like code or data section) or they are delay loaded only when we need them ?
If so can we unload these resources after we have used them in order to free allocated memory?
These functions are old, they date back to Windows versions that did not yet support virtual memory. Back in the olden days they would actually physically load a resource into RAM.
Those days are long gone, the OS loader creates a memory-mapped file to map the executable file into memory. And anything from the file (code and resources) are only mapped into RAM when the program dereferences a pointer. You only pay for what you use.
So LoadResource() does very little, it simply returns a pointer, disguised as a HGLOBAL handle. LockResource() does nothing interesting, it simply casts the HGLOBAL back to a pointer. When you actually start using it then you'll trip a page fault and the kernel reads the file, loading it into RAM. UnlockResource() and FreeResource() do nothing. If the OS needs RAM for another process then it can unmap the RAM for the resource. Nothing needs to be preserved since the memory is backed by the file, the page can simply be discarded. Paged back in when necessary if you use the resource again.

Why does the memory mapped file ever need to be flushed when access is RDWR?

I was reading through one of golang's implementation of memory mapped files, https://github.com/edsrzf/mmap-go/. First he describes the several access modes:
// RDONLY maps the memory read-only.
// Attempts to write to the MMap object will result in undefined behavior.
RDONLY = 0
// RDWR maps the memory as read-write. Writes to the MMap object will update the
// underlying file.
RDWR = 1 << iota
// COPY maps the memory as copy-on-write. Writes to the MMap object will affect
// memory, but the underlying file will remain unchanged.
COPY
But in gommap test file I see this:
func TestReadWrite(t *testing.T) {
mmap, err := Map(f, RDWR, 0)
... omitted for brevity...
mmap[9] = 'X'
mmap.Flush()
So why does he need to call Flush to make sure the contents are written to the file if the access mode is RDWR?
Or is the OS managing this so it only writes when it thinks it should?
If the last option, could you please explain it in a little more detail - what i read is that when the OS is low in memory it writes to the file and frees up memory. Is this correct and does it apply only to RDWR or only to COPY?
Thanks
The program maps a region of memory using mmap. It then modifies the mapped region. The system isn't required to write those modifications back to the underlying file immediately, so a read call on that file (in ioutil.ReadAll) could return the prior contents of the file.
The system will write the changes to the file at some point after you make the changes.
It is allowed to write the changes to the file any time after the changes are made, but by default makes no guarantees about when it writes those changes. All you know is that (unless the system crashes), the changes will be written at some point in the future.
If you need to guarantee that the changes have been written to the file at some point in time, then you must call msync.
The mmap.Flush function calls msync with the MS_SYNC flag. When that system call returns, the system has written the modifications to the underlying file, so that any subsequent call to read will read the modified file.
The COPY option sets the mapping to MAP_PRIVATE, so your changes will never be written back to the file, even if you using msync (via the Flush function).
Read the POSIX documentation about mmap and msync for full details.

Doing a zero-copy move of data from a Linux kernel buffer to hard disk

am trying to move data from a buffer in kernel space into the hard
disk without having to incur any additional copies from kernel buffer to
user buffers or any other kernel buffers. Any ideas/suggestions would be
most helpful.
The use case is basically a demux driver which collects data into a
demux buffer in kernel space and this buffer has to be emptied
periodically by copying the contents into a FUSE-based partition on the
disk. As the buffer gets full, a user process is signalled which then
determines the sector numbers on the disk the contents need to be copied
to.
I was hoping to mmap the above demux kernel buffer into user address
space and issue a write system call to the raw partition device. But
from what I can see, the this data is being cached by the kernel on its
way to the Hard Disk driver. And so I am assuming that involves
additional copies by the linux kernel.
At this point I am wondering if there is any other mechansim to do this
without involving additional copies by the kernel. I realize this is an
unsual usage scenario for non-embedded environments, but I would
appreciate any feedback on possible options.
BTW - I have tried using O_DIRECT when opening the raw partition, but
the subsequent write call fails if the buffer being passed is the
mmapped buffer.
Thanx!
You need to expose your demux buffer as a file descriptor (presumably, if you're using mmap() then you're already doing this - great!).
On the kernel side, you then need to implement the splice_read member of struct file_operations.
On the userspace side, create a pipe(), then use splice() twice - once to move the data from the demux file descriptor into the pipe, and a second time to move the data from the pipe to the disk file. Use the SPLICE_F_MOVE flag.
As documented in the splice() man page, it will avoid actual copies where it can, by copying references to pages of kernel memory rather than the pages themselves.

How do they read clusters/cylinders/sectors from the disk?

I needed to recover the partition table I deleted accidentally. I used an application named TestDisk. Its simply mind blowing. I reads each cylinder from the disk. I've seen similar such applications which work with MBR & partitioning.
I'm curious.
How do they read
clusters/cylinders/sectors from the
disk? Is there some kind of API for this?
Is it again OS dependent? If so whats the way to for Linux & for windows?
EDIT:
Well, I'm not just curious I want a hands on experience. I want to write a simple application which displays each LBA.
Cylinders and sectors (wiki explanation) are largely obsoleted by the newer LBA (logical block addressing) scheme for addressing drives.
If you're curious about the history, use the Wikipedia article as a starting point. If you're just wondering how it works now, code is expected to simply use the LBA address (which works largely the same way as a file does - a linear array of bytes arranged in blocks)
It's easy due to the magic of *nix special device files. You can open and read /dev/sda the same way you'd read any other file.
Just use open, lseek, read, write (or pread, pwrite). If you want to make sure you're physically fetching data from a drive and not from kernel buffers you can open with the flag O_DIRECT (though you must perform aligned reads/writes of 512 byte chunks for this to work).
For *nix, there have been already answers (/dev directory); for Windows, there are the special objects \\.\PhisicalDriveX, with X as the number of the drive, which can be opened using the normal CreateFile API. To actually perform reads or writes you have then to use the DeviceIoControl function.
More info can be found in "Physical Disks and Volumes" section of the CreateFile API documentation.
I'm the OP. I'm combining Eric Seppanen's & Matteo Italia's answers to make it complete.
*NIX Platforms:
It's easy due to the magic of *nix special device files. You can open and read /dev/sda the same way you'd read any other file.
Just use open, lseek, read, write (or pread, pwrite). If you want to make sure you're physically fetching data from a drive and not from kernel buffers you can open with the flag O_DIRECT (though you must perform aligned reads/writes of 512 byte chunks for this to work).
Windows Platform
For Windows, there are the special objects \\.\PhisicalDriveX, with X as the number of the drive, which can be opened using the normal CreateFile API. To perform reads or writes simply call ReadFile and WriteFile (buffer must be aligned on sector size).
More info can be found in "Physical Disks and Volumes" section of the CreateFile API documentation.
Alternatively you can also you DeviceIoControl function which sends a control code directly to a specified device driver, causing the corresponding device to perform the corresponding operation.
On linux, as root, you can save your MBR like this (Assuming you drive is /dev/sda):
dd if=/dev/sda of=mbr bs=512 count=1
If you wanted to read 1Mb from you drive, starting at the 10th MB:
dd if=/dev/sda of=1Mb bs=1Mb count=1 skip=10

Resources