This version of this document is no longer maintained. For the latest documentation, see http://www.qnx.com/developers/docs.

`devf-ram`

Simulate flash filesystem using RAM memory

You must be root to start this driver.

Syntax:

devf-ram
    [-a] [-b priority] 
    [-E] [-f verifylevel] [-i arrayindex[,partindex]] 
    [-l] [-m mountover]
    [-p backgroundpercent[,superlimit]] [-r] 
    [-s base[,wsize[,aoffset[,asize[,usize[,bwidth[,ileave]]]]]]]
    [-t threads] [-u update] [-V] [-v]
    [-w buffersize]

Options:

-a

Don't automount filesystem partitions present on the media. See also the -R option.

-b priority

Enable background reclaim at the specified priority. By default, background reclamation is disabled.

-E

Do not daemonize. Exit on EBADFSYS with partition num + 1 of corrupt filesystem.

-f verifylevel

Simulate flash verify; only provided for syntax compatibity with real flash hardware (default=0, 0=none, write=1, erase=2, all=3).

-i arrayindex[,partindex]

Starting socket index and first partition index; 0 >= index >=15. The default is 0,0. Use this to give multiple drivers unique IDs. The -i option is just a suggestion for the resource database manager; the selected indexes can be larger.

-l

List the available flash databases and exit.

-m mountover

Override the mountpoints assigned to a file system that are formatted with an empty (i.e. flashctl -p/dev/fs0p0 -e -f -n "") mountpoint. The mountover argument can include two %X format specifiers (like those for printf()) that are replaced by the socket index and the partition index.

The -m option doesn't override a mountpoint specified with mkefs.

-p backgroundpercent[,superlimit]

Set the background-reclaim percentage trigger (stale space over free space) and, optionally, the superseded extent limit before reclaim. The default is 100,16.

-R

Mount the automount filesystem as read-only. It doesn't affect raw partition mounts. The -R option has an effect only at startup and initialization. Any subsequent mounting (with either flashctl or mount) ignores the -R option. If you use -a option, the -R option is ignored.

-r

Specify this option always. Enable fault recovery for dirty extents, dangling extents, and partial reclaims. If you don't specify -r, recovery isn't performed, which, when power faults are likely, can waste space on the media or make the media read-only. You should always specify the -r option unless you're trying to debug a flash corruption issue or you don't have the time to repair a damage.

-s base[,wsize[,aoffset[,asize[,usize[,bwidth[,ileave]]]]]]

Set socket options, normally the base physical address, window size, array offset, array size, unit size, bus width, and interleave. The format is left flexible for socket services with customized drivers.

The arguments are:

base

Physical base address of the flash part. This value is board-specific.

For the devf-ram utility, the base argument carries a special meaning:

0: Allocate system memory.
Nonzero: Use the exact physical address. You must exercise caution here. See the caveats below.

wsize

Size of the physically contiguous flash part.

aoffset

For SRAM, the offset from the base address to the start of the flash array.

asize

For SRAM, the size of the flash array. The default is equal to wsize.

usize

The size of a physical erase sector. For SRAM, this number can be any power of two. 64K should be the minimum, for performance reasons.

bwidth

The total width of the data bus, as seen from the microprocessor's perspective. This is the width of one simulated flash chip multiplied by the interleave. The value is specified as a power of 2 (1,2,4,8).

ileave

The number of simulated flash chips arranged on the data bus.

You can specify the base physical address, sizes, and offset in octal (0777), hexadecimal (0x1ff), or decimal (511). The sizes must be a power of two, and you can specify them with any of the following suffixes:

(nothing) -- bytes
k -- kilobytes
m -- megabytes

On ARM targets, devf-ram can't resize the shared object /dev/shmem/fs*. If you need to restart devf-ram with a new size, first unlink the old shared object:

rm /dev/shmem/fs*

-t threads

Number of threads; 1 >= threads >= 4 (default is 2). Extra threads increase performance when background reclaim is enabled (with the -b option) and when multiple chips and/or spare blocks are available.

-u update

Update level for timestamps; 0 for never update, 1 to update files only, and 2 to update files and directories. The default update is 0.

-V

Display filesystem and MTD version information.

-v

Display verbose information.

-w buffersize

Write (append) buffer size in bytes. The default buffersize is 512. Using a larger write-buffer prevents the creation of very small extents, reducing overhead. If buffersize is 0, appending is disabled.

Description:

The devf-ram manager simulates flash filesystem in RAM using the following default filenames (the ID, n, appended to /dev/fs can be changed via the -i option):

/dev/fsn: Default mountpoint for socket n.
/dev/fsnp0: Raw access for socket n, partition 0.
mountpoint: Flash filesystem mountpoint for socket n, partition 0 with transparent decompression.

You should always specify the -r option unless you're trying to debug a flash corruption issue or you don't have the time to repair a damage. See the background information for this:

If an erase was happening when the power is cut off, it results in a number of dangling extents at the next power on. These extents continue to occupy space forever, until they are deleted. Using the -r option will cause them to be deleted. If you start the the driver with -vv, it prints dangle for every dangling extent found.
If the filesystem detects an error, and the -r isn't specified, the driver marks the partition read-only, so that more damage isn't done. The second case is when a reclaim was interrupted by a power-loss. In this case the spare block may be unusable and the driver prints partial to the console. The partition is still read-write, but reclaims are turned off, which means overwriting of files will eventually fill up the filesystem with stale data.

You can specify the mountpoint above with the mount attribute of the mkefs command, and override it with the -n option to flashctl. By default, it's /fsnp0.

Examples:

Start devf-ram with a 16MB partition.

devf-ram  -s0,16m

Start devf-ram and automatically mount the flash filesystem partitions, with an initial fault recovery process, most POSIX semantics enabled and background reclaim at priority 5 (default size: 1M):

devf-ram -r -u2 -b5 &

Create a 32MB flash partition, allocated from system RAM, with a 64KB unit (sector) size:

devf-ram -s0,32m,,,64k  -v -r

Create a 128MB flash partition in system RAM, with large block sizes (to speed formatting):

devf-ram -s0,128m,,,512k -v -r

Create a 4MB partition from system RAM:

devf-ram -s0,4m,,,64k -v -r

You must format and erase a devf-ram partition before you can mount the flash filesystem. See the caveats below.

If you specify a blocksize in your buildfile for DRAM-based flash filesystems, limit the size to the default, which is 64K.

Caveats:

Although the flash filesystem supports most POSIX semantics, some functionality isn't implemented in order to keep the driver simple and efficient. The unsupported POSIX semantics include:

Hard links, and everything related to hard links (the . and .. directories don't exist, struct stat's nlink member is hard-coded, and unlink() of directories returns ENOTSUP).
Access times aren't updated on the media; they're set to the modification time.

QNX Neutrino flash filesystem version 3 no longer provides built-in decompression. The flash filesystem's decompression functionality has moved into the inflator resource manager. You should now use the deflate utility to compress files.

Performance might be slow when multiple writers are writing randomly to a shared file or to a shared directory (e.g. using unlink or rename). In these cases, the offset pointers have to be rewound for every access. There's no performance penalty when appending to a file, or when creating files with open(O_CREAT), mkdir, mknod, or link.

Don't try to create a devf-ram partition at the address of a real flash memory. You may get an error message: Unable to properly identify any flash devices.

Don't try to create a devf-ram partition (e.g. using nonzero value for base argument) at the address of physical memory that is in use. It may destroy applications and crash the operating system. The only use for specifying such nonzero base is to create a flash filesystem for board specific memory (e.g. SRAM).

You must format and erase a devf-ram partition before you can mount the flash filesystem. e.g.

devf-ram -s0,16m
flashctl -p /dev/fs0p0 -e -f -m

If there's insufficient RAM, when you try to create an nM size partition with -s0 option, the devf-ram driver returns without an error message. The partition isn't created.