kernel/linux.git/drivers/tty/n_tty.c, branch v6.6.132

tty: n_tty: use uint for space returned by tty_write_room()

2025-04-10T12:37:34+00:00

[ Upstream commit d97aa066678bd1e2951ee93db9690835dfe57ab6 ] tty_write_room() returns an "unsigned int". So in case some insane driver (like my tty test driver) returns (legitimate) UINT_MAX from its tty_operations::write_room(), n_tty is confused on several places. For example, in process_output_block(), the result of tty_write_room() is stored into (signed) "int". So this UINT_MAX suddenly becomes -1. And that is extended to ssize_t and returned from process_output_block(). This causes a write() to such a node to receive -EPERM (which is -1). Fix that by using proper "unsigned int" and proper "== 0" test. And return 0 constant directly in that "if", so that it is immediately clear what is returned ("space" equals to 0 at that point). Similarly for process_output() and __process_echoes(). Note this does not fix any in-tree driver as of now. If you want "Fixes: something", it would be commit 03b3b1a2405c ("tty: make tty_operations::write_room return uint"). I intentionally do not mark this patch by a real tag below. Signed-off-by: Jiri Slaby (SUSE) Link: https://lore.kernel.org/r/20250317070046.24386-6-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman Signed-off-by: Sasha Levin

tty: n_tty: Fix buffer offsets when lookahead is used

2024-06-21T12:38:24+00:00

commit b19ab7ee2c4c1ec5f27c18413c3ab63907f7d55c upstream. When lookahead has "consumed" some characters (la_count > 0), n_tty_receive_buf_standard() and n_tty_receive_buf_closing() for characters beyond the la_count are given wrong cp/fp offsets which leads to duplicating and losing some characters. If la_count > 0, correct buffer pointers and make count consistent too (the latter is not strictly necessary to fix the issue but seems more logical to adjust all variables immediately to keep state consistent). Reported-by: Vadym Krevs Fixes: 6bb6fa6908eb ("tty: Implement lookahead to process XON/XOFF timely") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218834 Tested-by: Vadym Krevs Cc: stable@vger.kernel.org Signed-off-by: Ilpo Järvinen Link: https://lore.kernel.org/r/20240514140429.12087-1-ilpo.jarvinen@linux.intel.com Signed-off-by: Greg Kroah-Hartman

tty: n_tty: deduplicate copy code in n_tty_receive_buf_real_raw()

2023-08-27T09:46:52+00:00

The code is duplicated to perform the copy twice -- to handle buffer wrap-around. Instead of the duplication, roll this into the loop. (And add some blank lines around to have the code a bit more readable.) Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-15-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: extract ECHO_OP processing to a separate function

2023-08-27T09:46:52+00:00

__process_echoes() contains ECHO_OPs processing. It is stuffed in a while loop and the whole function is barely readable. Separate it to a new function: n_tty_process_echo_ops(). Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-14-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: unify counts to size_t

2023-08-27T09:46:52+00:00

Some count types are already 'size_t' for a long time. Some were switched to 'size_t' recently. Unify the rest with those now. This allows for some min_t()s to become min()s. And make one min() an explicit min_t() as we are comparing signed 'room' to unsigned 'count'. Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-13-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: use u8 for chars and flags

2023-08-27T09:46:52+00:00

Unify with the tty layer and use u8 for both chars and flags. Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-12-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: simplify chars_in_buffer()

2023-08-27T09:46:52+00:00

The 'if' in chars_in_buffer() is misleadingly inverted. And since the only difference is the head used for computation, cache the head using ternary operator. And use that in return directly. Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-11-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: remove unsigned char casts from character constants

2023-08-27T09:46:52+00:00

We compile with -funsigned-char, so all character constants are already unsigned chars. Therefore, remove superfluous casts. Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-10-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: move newline handling to a separate function

2023-08-27T09:46:52+00:00

Currently, n_tty handles the newline in a label in n_tty_receive_char_canon(). That is invoked from two more places. Split this code to a separate function and avoid the label in this case. This makes the code flow more understandable. Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-9-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman

tty: n_tty: move canon handling to a separate function

2023-08-27T09:46:52+00:00

n_tty_receive_char_special() is already complicated enough. Split the canon handling to a separate function: n_tty_receive_char_canon(). Signed-off-by: "Jiri Slaby (SUSE)" Link: https://lore.kernel.org/r/20230827074147.2287-8-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman