Add a new plugin for creating pipes between pairs of VFS handles. It is
intended to replace the libc_pipe plugin, one of the last remaining libc
plugins.
In contrast to the libc_pipe plugin, this plugin defers cross-handle
notification until I/O signal handling rather than block and unblock
readers using a semaphore. This is a performance regression in the case
of multiple threads blocking on a pipe, but shall be an intermediate
mechanism pending renovations within the libc VFS and threading layers.
As a side effect, threads blocked on a pipe might not be resumed until
the main thread suspends and dispatches I/O signals.
The "test-libc_pipe" test has been adjusted to use the VFS pipe plugin
and tests both local pipes and pipes hosted remotely in the VFS server.
Merge adaptations (such as EOF handling, adjustment to VFS/libc
interface changes) by Norman Feske.
Fix#2303
This patch improves the libc's write operation to iterate on partial
writes to continuous files until the original write count is reached.
The split of large write operations into small partial writes as
dictated by the VFS infrastructure (e.g., constained by I/O buffer
sizes) becomes invisible to the libc-using application.
Issue #3507
Issue #2303
This patch introduces a new scheme of handling ioctl operations that
maps ioctls to pseudo-file accesses, similar to how the libc maps socket
calls to socket-fs operations.
A device file can be accompanied with a (hidden) directory that is named
after the device file and hosts pseudo files for triggering the various
device operations. For example, for accessing a terminal, the directory
structure looks like this:
/dev/terminal
/dev/.terminal/info
The 'info' file contains device information in XML format. The type of
the XML node corresponds to the device type. E.g., If the libc receives
a 'TIOCGWINSZ' ioctl for /dev/terminal, it reads the content of
/dev/.terminal/info to obtain the terminal-size information. In this
case, the 'info' file looks as follows:
<terminal rows="25" columns="80/>
Following this scheme, VFS plugins can support ioctl operations by
providing an ioctl directory in addition to the actual device file.
Internally, the mechanism uses the 'os/vfs.h' API to access pseudo
files. Hence, we need to propagate the Vfs::Env to 'vfs_plugin.cc' to
create an instance of a 'Directory' for the root for the VFS.
Issue #3519
This patch extends the 'File_system::Status',
'File_system::Directory_entry', and the related 'Vfs' types with
the following additional information:
- Distinction between continuous and transactional files (Node_type)
(issue #3507)
- Readable, writeable, and executable attributes (Node_rwx),
replacing the former 'mode' bits
(issue #3030)
The types 'Node_rwx', 'Node_type' are defined twice,
once for the VFS (vfs/types.h) and once for the 'File_system'
session (file_system_session/file_system_session.h).
Similarly, there is a direct correspondance between
'Vfs::Directory_service::Dirent' and 'File_system::Directory_entry'.
This duplication of types follows the existing pattern of keeping the
VFS and file-system session independent from each other.
By specifying <libc update_mtime="no"...>, the modification-time update
on VFS-sync operations (as issued whenever a written file is closed)
can explicitly be disabled.
Issue #1784
This patch replaces the naive dup2 implementation (that merely
duplicated the context pointer) by the replication of the original
FD state by re-opening the same file with the same flags and seek
position. This prevents a potential double release of the VFS handle
(the FD context). It also implements 'dup'.
Fixes#3505Fixes#3477
This patch unifies the patterns of using the 'Genode' and 'Libc'
namespaces.
Types defined in the 'internal/' headers reside in the 'Libc'
namespace. The code in the headers does not need to use the
'Libc::' prefix.
Compilation units import the 'Libc' namespace after the definition of
local types. Local types reside in the 'Libc' namespace (and should
eventually move to an 'internal/' header).
Since the 'Libc' namespace imports the 'Genode' namespace, there is
no need to use the 'Genode::' prefix. Consequently, code in the
compilation units rarely need to qualify the 'Genode' or 'Libc'
namespaces.
There are a few cases where the 'Libc', the 'Genode', and the global
(libc) namespaces are ambigious. In these cases, an explicit
clarification is needed:
- 'Genode::Allocator' differs from 'Libc::Allocator'.
- 'Genode::Env' differs from 'Libc::Env'.
- Genode's string functions (strcmp, memcpy, strcpy) conflict
with the names of the (global) libc functions.
- There exist both 'Genode::uint64_t' and the libc'c 'uint64_t'.
Issue #3497
This patch is the first step of re-organizing the internal structure of
the libc. The original version involved many direct calls of global
functions (often with side effects) across compilation units, which
made the control flow (e.g., the initialization sequence) hard to
follow.
The new version replaces those ad-hoc interactions with dedicated
interfaces (like suspend.h, resume.h, select.h, current_time.h). The
underlying facilities are provided by the central Libc::Kernel and
selectively propagated to the various compilation units. The latter is
done by a sequence of 'init_*' calls, which eventually will be replaced
by constructor calls.
The addition of new headers increases the chance for name clashes with
existing (public) headers. To disambiguate libc-internal header files
from public headers, this patch moves the former into a new 'internal/'
subdirectory. This makes the include directives easier to follow and the
libc's source-tree structure more tidy.
There are still a few legacies left, which cannot easily be removed
right now (e.g., because noux relies on them). However, the patch moves
those bad apples to legacy.h and legacy.cc, which highlights the
deprecation of those functions.
Issue #3497
Move the _vfs_sync implementation from vfs_plugin.h to vfs_plugin.cc and
wrap all VFS access in the VFS_THREAD_SAFE macro.
Syncing must be made thread safe because it often happens when closing
files and may require suspending the current thread for I/O signal
dispatching.
Ref #3409
Replace the I/O response handler that is passed to the VFS at
construction with an object that is dynamically attached to handles.
This object shall also accept read-ready notifications, and plugins are
encouraged to keep handles awaiting ready-ready notifications separate
from handles that await I/O progress.
Replace the use of handle lists in plugins with handle queues, this
makes the code easier to understand and the ordering of notifications to
the application more explicit.
These changes replace the use of the Post_signal_hook from all VFS
plugins, applications must assume that read-ready and I/O notifications
occur during I/O signal dispatch and use an Io_progress_handler at its
entrypoints to defer response until after signal dispatching.
Fix#3257
When 'notify_read_ready()' is called during 'select()' and fails,
suspending can cause a deadlock when the libc IO response handler becomes
active and calls 'select_notify()', which tries to acquire the
'select callback list lock', which is already acquired by the suspended
'select()' call.
It seems possible to ignore a failed 'notify_read_ready()' call instead of
suspending. When the VFS plugin calls the IO handler later when the
notification request can be processed, the 'select_notify()' call of the
libc IO response handler will eventually call 'notify_read_ready()' again.
Fixes#2970
Calling 'handle_io_response()' in a regular VFS function (in contrast to a
post-signal hook) can cause problems if the caller of the VFS function holds
a lock which prevents the io response handler from returning.
With this commit, the user of the VFS becomes responsible for unblocking
threads which might be blocking after a failed 'queue_read()', 'queue_sync()'
or 'write()' call.
Fixes#2896
The libc will now use the file given by the 'nameserver_file' attribute
to get the DNS nameserver address instead of reading '/etc/resolv.conf'.
It defaults to '/socket/nameserver' which is the common location when
using the lxip VFS plugin. As a constraint the libc will read the first
line and expects the nameserver address without any keywords in front of
it.
Fixes#2861.
The VFS library can be used in single-threaded or multi-threaded
environments and depending on that, signals are handled by the same thread
which uses the VFS library or possibly by a different thread. If a VFS
plugin needs to block to wait for a signal, there is currently no way
which works reliably in both environments.
For this reason, this commit makes the interface of the VFS library
nonblocking, similar to the File_system session interface.
The most important changes are:
- Directories are created and opened with the 'opendir()' function and the
directory entries are read with the recently introduced 'queue_read()'
and 'complete_read()' functions.
- Symbolic links are created and opened with the 'openlink()' function and
the link target is read with the 'queue_read()' and 'complete_read()'
functions and written with the 'write()' function.
- The 'write()' function does not wait for signals anymore. This can have
the effect that data written by a VFS library user has not been
processed by a file system server yet when the library user asks for the
size of the file or closes it (both done with RPC functions at the file
system server). For this reason, a user of the VFS library should
request synchronization before calling 'stat()' or 'close()'. To make
sure that a file system server has processed all write request packets
which a client submitted before the synchronization request,
synchronization is now requested at the file system server with a
synchronization packet instead of an RPC function. Because of this
change, the synchronization interface of the VFS library is now split
into 'queue_sync()' and 'complete_sync()' functions.
Fixes#2399
The socket file system can be configured in the "socket" attribute of
the libc config node like follows.
<vfs> <dir name="socket"> <fs/> </dir> </vfs>
<libc ... socket="/socket"/>
This configures the socket file system libc backend to access files in
"/socket" for socket operations.
The support has two parts. First, a VFS plugin now gets passed an
I/O-response handler callback on construction, which informs users of the
VFS that an I/O event occurred. This enables, for example, the libC to
check if blocking read can be completed. Further, the VFS file I/O
interface provides now functions for suspendable reads, i.e.,
queue_read() and complete_read().
Christian Prochaska
Ehmry -
Norman Feske
Christian Helmuth
Josef Söntgen
Alexander Senier
Ben Larson
Sebastian Sumpf
Emery Hemingway
Emery Hemingway
Alexander Boettcher