On the embedded platform, certain critical data, such as IMU data, is
transmitted through UART. The tty_flip_buffer_push() interface in the TTY
layer uses system_dfl_wq to handle the flipping of the TTY buffer.
Although the unbound workqueue can create new threads on demand and wake
up the kworker thread on an idle CPU, it may be preempted by real-time
tasks or other high-prio tasks.
flush_to_ldisc() needs to wake up the relevant data handle thread. When
executing __wake_up_common_lock(), it calls spin_lock_irqsave(), which
does not disable preemption but disables migration in RT-Linux. This
prevents the kworker thread from being migrated to other cores by CPU's
balancing logic, resulting in long delays. The call trace is as follows:
__wake_up_common_lock
__wake_up
ep_poll_callback
__wake_up_common
__wake_up_common_lock
__wake_up
n_tty_receive_buf_common
n_tty_receive_buf2
tty_ldisc_receive_buf
tty_port_default_receive_buf
flush_to_ldisc
In our system, the processing interval for each frame of IMU data
transmitted via UART can experience significant jitter due to this issue.
Instead of the expected 10 to 15 ms frame processing interval, we see
spikes up to 30 to 35 ms. Moreover, in just one or two hours, there can
be 2 to 3 occurrences of such high jitter, which is quite frequent. This
jitter exceeds the software's tolerable limit of 20 ms.
Introduce flip_wq in tty_port which can be set by tty_port_link_wq() or as
default linked to default workqueue allocated when tty_register_driver().
The default workqueue is allocated with flag WQ_SYSFS, so that cpumask and
nice can be set dynamically. The execution timing of tty_port_link_wq() is
not clearly restricted. The newly added function tty_port_link_driver_wq()
checks whether the flip_wq of the tty_port has already been assigned when
linking the default tty_driver's workqueue to the port. After the user has
set a custom workqueue for a certain tty_port using tty_port_link_wq(), the
system will only use this custom workqueue, even if tty_driver does not
have %TTY_DRIVER_NO_WORKQUEUE flag. When tty_port register device, flip_wq
link operation is done by tty_port_link_driver_wq(), but for in-memory
devices the link operation cannot cover all the cases. Although
tty_port_install() is dedicated for in-memory devices lik PTY to link port
allocated on demand, the logic of tty_port_install() is so simple that
people may not call it, vc_cons[0].d->port is one such case. We check the
buf.flip_wq when flip TTY buffer, if buf.flip_wq of TTY port is NULL, use
system_dfl_wq as a backup.
To avoid naming conflict of the default tty_driver's workqueue, using
'"%s-%s", driver->name, driver->driver_name' as the workqueue name. In
cases where driver_name is not specified and therefore is NULL, the
workqueue is not created. Drivers that do not define driver_name are
potentially in-memory devices like vty, which generally do not require
special workqueue settings. Even with the combination of name and
driver_name, the workqueue names can still be duplicated, as many tty
serial drivers use "ttyS" as dev_name and "serial" as driver_name. I
modified the conflicting driver_name of these drivers by appending a
suffix of _xx based on the corresponding .c file. If this modification is
not made, it could not only lead to duplicate workqueue names but also
result in duplicate entries for the /proc/tty/driver/<driver_name> nodes.
Introduce %TTY_DRIVER_NO_WORKQUEUE flag meaning not to create the
default single tty_driver workqueue. Two reasons why need to introduce the
%TTY_DRIVER_NO_WORKQUEUE flag:
1. If the WQ_SYSFS parameter is enabled, workqueue_sysfs_register() will
fail when trying to create a workqueue with the same name. The pty is an
example of this; if both CONFIG_LEGACY_PTYS and CONFIG_UNIX98_PTYS are
enabled, the call to tty_register_driver() in unix98_pty_init() will fail.
2. Different TTY ports may be used for different tasks, which may require
separate core binding control via workqueues. In this case, the workqueue
created by default in the TTY driver is unnecessary. Enabling this flag
prevents the creation of this redundant workqueue.
After applying this patch, we can set the related UART TTY flip buffer
workqueue by sysfs. We set the cpumask to CPU cores associated with the
IMU tasks, and set the nice to -20. Testing has shown significant
improvement in the previously described issue, with almost no stuttering
occurring anymore.
Tested-by: Tommaso Merciai <tommaso.merciai.xr@bp.renesas.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Xin Zhao <jackzxcui1989@163.com>
Link: https://patch.msgid.link/20260213085039.3274704-1-jackzxcui1989@163.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is the result of running the Coccinelle script from
scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to
avoid scalar types (which need careful case-by-case checking), and
instead replace kmalloc-family calls that allocate struct or union
object instances:
Single allocations: kmalloc(sizeof(TYPE), ...)
are replaced with: kmalloc_obj(TYPE, ...)
Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...)
are replaced with: kmalloc_objs(TYPE, COUNT, ...)
Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...)
are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...)
(where TYPE may also be *VAR)
The resulting allocations no longer return "void *", instead returning
"TYPE *".
Signed-off-by: Kees Cook <kees@kernel.org>
Currently if a user enqueue a work item using schedule_delayed_work() the
used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use
WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to
schedule_work() that is using system_wq and queue_work(), that makes use
again of WORK_CPU_UNBOUND.
This lack of consistentcy cannot be addressed without refactoring the API.
This patch continues the effort to refactor worqueue APIs, which has begun
with the change introducing new workqueues and a new alloc_workqueue flag:
commit 128ea9f6cc ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566 ("workqueue: Add new WQ_PERCPU flag")
system_dfl_wq should be the default workqueue so as not to enforce
locality constraints for random work whenever it's not required.
The old system_unbound_wq will be kept for a few release cycles.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20251104105446.110884-1-marco.crivellari@suse.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The same as in the previous patch, tty_prepare_flip_string() accepts
size_t as an size argument. It returns the same size (or less). It is
unexpected that it returns a signed value and can confuse users to check
for negative values.
Instead, return the same size_t as accepted to make clear we return
values >= 0, where zero in fact means failure.
Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230816105530.3335-7-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
All the functions accept size_t as a size argument. They finally return
the same size (or less). It is quite unexpected that they return a
signed value and can confuse users to check for negative values.
Instead, return the same size_t as accepted to make clear we return
values >= 0, where zero in fact means failure.
Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230816105530.3335-6-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It comes from both paste_selection() and tty_port_default_receive_buf()
as unsigned (int and size_t respectively). Switch to size_t to converge
to that eventually.
Return the count as size_t too (the two callers above expect that).
Switch paste_selection()'s type of 'count' too, so that the returned and
passed type match.
Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230810091510.13006-13-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The struct tty_buffer has flags which is only used for storing TTYB_NORMAL.
There is also a few quite confusing operations for checking the presense
of TTYB_NORMAL. Simplify things by converting flags to bool.
Despite the name remaining the same, the meaning of "flags" is altered
slightly by this change. Previously it referred to flags of the buffer
(only TTYB_NORMAL being used as a flag). After this change, flags tell
whether the buffer contains/should be allocated with flags array along
with character data array. It is much more suitable name that
TTYB_NORMAL was for this purpose, thus the name remains.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20221019105504.16800-1-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Pull tty / serial driver updates from Greg KH:
"Here is the big set of tty and serial driver changes for 6.0-rc1.
It was delayed from last week as I wanted to make sure the last commit
here got some good testing in linux-next and elsewhere as it seemed to
show up only late in testing for some reason.
Nothing major here, just lots of cleanups from Jiri and Ilpo to make
the tty core cleaner (Jiri) and the rs485 code simpler to use (Ilpo).
Also included in here is the obligatory n_gsm updates from Daniel
Starke and lots of tiny driver updates and minor fixes and tweaks for
other smaller serial drivers.
All of these have been in linux-next for a while with no reported
problems"
* tag 'tty-6.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty: (186 commits)
tty: serial: qcom-geni-serial: Fix %lu -> %u in print statements
tty: amiserial: Fix comment typo
tty: serial: document uart_get_console()
tty: serial: serial_core, reformat kernel-doc for functions
Documentation: serial: link uart_ops properly
Documentation: serial: move GPIO kernel-doc to the functions
Documentation: serial: dedup kernel-doc for uart functions
Documentation: serial: move uart_ops documentation to the struct
dt-bindings: serial: snps-dw-apb-uart: Document Rockchip RV1126
serial: mvebu-uart: uart2 error bits clearing
tty: serial: fsl_lpuart: correct the count of break characters
serial: stm32: make info structs static to avoid sparse warnings
serial: fsl_lpuart: zero out parity bit in CS7 mode
tty: serial: qcom-geni-serial: Fix get_clk_div_rate() which otherwise could return a sub-optimal clock rate.
serial: 8250_bcm2835aux: Add missing clk_disable_unprepare()
tty: vt: initialize unicode screen buffer
serial: remove VR41XX serial driver
serial: 8250: lpc18xx: Remove redundant sanity check for RS485 flags
serial: 8250_dwlib: remove redundant sanity check for RS485 flags
dt_bindings: rs485: Correct delay values
...
There is a race in pty_write(). pty_write() can be called in parallel
with e.g. ioctl(TIOCSTI) or ioctl(TCXONC) which also inserts chars to
the buffer. Provided, tty_flip_buffer_push() in pty_write() is called
outside the lock, it can commit inconsistent tail. This can lead to out
of bounds writes and other issues. See the Link below.
To fix this, we have to introduce a new helper called
tty_insert_flip_string_and_push_buffer(). It does both
tty_insert_flip_string() and tty_flip_buffer_commit() under the port
lock. It also calls queue_work(), but outside the lock. See
71a174b39f (pty: do tty_flip_buffer_push without port->lock in
pty_write) for the reasons.
Keep the helper internal-only (in drivers' tty.h). It is not intended to
be used widely.
Link: https://seclists.org/oss-sec/2022/q2/155
Fixes: 71a174b39f (pty: do tty_flip_buffer_push without port->lock in pty_write)
Cc: 一只狗 <chennbnbnb@gmail.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Suggested-by: Hillf Danton <hdanton@sina.com>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20220707082558.9250-2-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
