- 24 Jul, 2013 8 commits
-
-
Peter Hurley authored
Since neither echo_commit nor echo_tail can change for the duration of __process_echoes loop, substitute index comparison for the snapshot counter. Signed-off-by:
Peter Hurley <peter@hurleysoftware.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Peter Hurley authored
Don't have the driver flush received echoes if no echoes were actually output. Signed-off-by:
-
Peter Hurley authored
Byte-by-byte echo output is painfully slow, requiring a lock/unlock cycle for every input byte. Instead, perform the echo output in blocks of 256 characters, and at least once per flip buffer receive. Enough space is reserved in the echo buffer to guarantee a full block can be saved without overrunning the echo output. Overrun is prevented by discarding the oldest echoes until enough space exists in the echo buffer to receive at least a full block of new echoes. Signed-off-by:
-
Peter Hurley authored
Use output_lock mutex as a memory barrier when storing echo_commit. Signed-off-by:
-
Peter Hurley authored
Adding data to echo_buf (via add_echo_byte()) is guaranteed to be single-threaded, since all callers are from the n_tty_receive_buf() path. Processing the echo_buf can be called from either the n_tty_receive_buf() path or the n_tty_write() path; however, these callers are already serialized by output_lock. Publish cumulative echo_head changes to echo_commit; process echo_buf from echo_tail to echo_commit; remove echo_lock. On echo_buf overrun, claim output_lock to serialize changes to echo_tail. Signed-off-by:
-
Peter Hurley authored
Prepare for lockless echo_buf handling; compute current byte count of echo_buf from head and tail indices. Signed-off-by:
-
Peter Hurley authored
Instead of using a single index to track the current echo_buf position, use a head index when adding to the buffer and a tail index when consuming from the buffer. Allow these head and tail indices to wrap at max representable value; perform modulo reduction via helper functions when accessing the buffer. Signed-off-by:
-
Peter Hurley authored
The echo_overrun field is only assigned and never tested; remove it. Signed-off-by:
-
- 23 Jul, 2013 32 commits
-
-
Peter Hurley authored
TTY_BUFFER_PAGE is only used within drivers/tty/tty_buffer.c; relocate to that file scope. Signed-off-by:
-
Peter Hurley authored
Convert the tty_buffer_flush() exclusion mechanism to a public interface - tty_buffer_lock/unlock_exclusive() - and use the interface to safely write the paste selection to the line discipline. Signed-off-by:
-
Peter Hurley authored
__tty_flush_buffer() is now only called by tty_flush_buffer(); merge functions. Signed-off-by:
-
Peter Hurley authored
Atomic bit ops are no longer required to indicate a flip buffer flush is pending, as the flush_mutex is sufficient barrier. Remove the unnecessary port .iflags field and localize flip buffer state to struct tty_bufhead. Signed-off-by:
-
Peter Hurley authored
Separate the head and tail ptrs to avoid cache-line contention (so called 'false-sharing') between concurrent threads. Signed-off-by:
-
Peter Hurley authored
Now that dropping the buffer lock is not necessary (as result of converting the spin lock to a mutex), the flip buffer flush no longer needs to be handled by the buffer work. Simply signal a flush is required; the buffer work will exit the i/o loop, which allows tty_buffer_flush() to proceed. Signed-off-by:
-
Peter Hurley authored
The buffer work may race with parallel tty_buffer_flush. Use a mutex to guarantee exclusive modify access to the head flip buffer. Remove the unneeded spin lock. Signed-off-by:
-
Peter Hurley authored
Driver-side flip buffer input is already single-threaded; 'publish' the .next link as the last operation on the tail buffer so the 'consumer' sees the already-completed flip buffer. The commit buffer index is already 'published' by driver-side functions. Signed-off-by:
-
Peter Hurley authored
Lockless flip buffers require atomically updating the bytes-in-use watermark. The pty driver also peeks at the watermark value to limit memory consumption to a much lower value than the default; query the watermark with new fn, tty_buffer_space_avail(). Signed-off-by:
-
Peter Hurley authored
Use a 0-sized sentinel to avoid assigning the head ptr from the driver side thread. This also eliminates testing head/tail for NULL. When the sentinel is first 'consumed' by the buffer work (or by tty_buffer_flush()), it is detached from the list but not freed nor added to the free list. Both buffer work and tty_buffer_flush() continue to preserve at least 1 flip buffer to which head & tail is pointed. Signed-off-by:
-
Peter Hurley authored
In preparation for lockless flip buffers, make the flip buffer free list lockless. NB: using llist is not the optimal solution, as the driver and buffer work may contend over the llist head unnecessarily. However, test measurements indicate this contention is low. Signed-off-by:
-
Peter Hurley authored
Signed-off-by:
-
Peter Hurley authored
tty_buffer_find() implements a simple free list lookaside cache. Merge this functionality into tty_buffer_alloc() to reflect the more traditional alloc/free symmetry. Signed-off-by:
-
Peter Hurley authored
Factor shared code; prepare for adding 0-sized sentinel flip buffer. Signed-off-by:
-
Peter Hurley authored
Since flip buffers are size-aligned to 256 bytes and all flip buffers 512-bytes or larger are not added to the free list, the free list only contains 256-byte flip buffers. Remove the list search when allocating a new flip buffer. Signed-off-by:
-
Peter Hurley authored
The char_buf_ptr and flag_buf_ptr values are trivially derived from the .data field offset; compute values as needed. Fixes a long-standing type-mismatch with the char and flag ptrs. Signed-off-by:
-
Peter Hurley authored
Scheduling buffer work on the same cpu as the read() thread limits the parallelism now possible between the receive_buf path and the n_tty_read() path. Signed-off-by:
-
Peter Hurley authored
The pty driver forces ldisc flow control on, regardless of available receive buffer space, so the writer can be woken whenever unthrottle is called. However, this 'forced throttle' has performance consequences, as multiple atomic operations are necessary to unthrottle and perform the write wakeup for every input line (in canonical mode). Instead, short-circuit the unthrottle if the tty is a pty and perform the write wakeup directly. Signed-off-by:
-
Peter Hurley authored
Prepare to special case pty flow control; avoid forward declaration. Signed-off-by:
-
Peter Hurley authored
Prepare for special handling of pty throttle/unthrottle; factor flow control into helper functions. Signed-off-by:
-
Peter Hurley authored
Prepare to factor throttle and unthrottle into helper functions; relocate chars_in_buffer() to avoid forward declaration. Signed-off-by:
-
Peter Hurley authored
No tty driver modifies termios during throttle() or unthrottle(). Therefore, only read safety is required. However, tty_throttle_safe and tty_unthrottle_safe must still be mutually exclusive; introduce throttle_mutex for that purpose. Signed-off-by:
-
Peter Hurley authored
If the read buffer indices are in the same cache-line, cpus will contended over the cache-line (so called 'false sharing'). Separate the producer-published fields from the consumer-published fields; document the locks relevant to each field. Signed-off-by:
-
Peter Hurley authored
User-space read() can run concurrently with receiving from device; waiting for receive_buf() to complete is not required. Signed-off-by:
-
Peter Hurley authored
Signed-off-by:
-
Peter Hurley authored
lnext escapes the next input character as a literal, and must be reset when canonical mode changes (to avoid misinterpreting a special character as a literal if canonical mode is changed back again). lnext is specifically not reset on a buffer flush so as to avoid misinterpreting the next input character as a special character. Signed-off-by:
-
Peter Hurley authored
n_tty has a single-producer/single-consumer input model; use lockless publish instead. Use termios_rwsem to exclude both consumer and producer while changing or resetting buffer indices, eg., when flushing. Also, claim exclusive termios_rwsem to safely retrieve the buffer indices from a thread other than consumer or producer (eg., TIOCINQ ioctl). Note the read_tail is published _after_ clearing the newline indicator in read_flags to avoid racing the producer. Drop read_lock spinlock. Signed-off-by:
-
Peter Hurley authored
canon_data represented the # of lines which had been copied to the receive buffer but not yet copied to the user buffer. The value was tested to determine if input was available in canonical mode (and also to force input overrun if the receive buffer was full but a newline had not been received). However, the actual count was irrelevent; only whether it was non-zero (meaning 'is there any input to transfer?'). This shared count is unnecessary and unsafe with a lockless algorithm. The same check is made by comparing canon_head with read_tail instead. Signed-off-by:
-
Peter Hurley authored
Use termios_rwsem to guarantee safe access to the termios values. This is particularly important for N_TTY as changing certain termios settings alters the mode of operation. termios_rwsem must be dropped across throttle/unthrottle since those functions claim the termios_rwsem exclusively (to guarantee safe access to the termios and for mutual exclusion). Signed-off-by:
-
Peter Hurley authored
termios is commonly accessed unsafely (especially by N_TTY) because the existing mutex forces exclusive access. Convert existing usage. Signed-off-by:
-
Peter Hurley authored
Storing the read_cnt creates an unnecessary shared variable between the single-producer (n_tty_receive_buf()) and the single-consumer (n_tty_read()). Compute read_cnt from head & tail instead of storing. Signed-off-by:
-
Peter Hurley authored
Wrap read_buf indices (read_head, read_tail, canon_head) at max representable value, instead of at the N_TTY_BUF_SIZE. This step is necessary to allow lockless reads of these shared variables (by updating the variables atomically). Signed-off-by:
-
