OrgApache/incubator-nuttx
OrgApache
/
incubator-nuttx
mirror of https://github.com/apache/incubator-nuttx.git
1
0
Fork 0
Code Issues Releases Wiki Activity
incubator-nuttx/mm/iob/Kconfig

120 lines
3.9 KiB
Plaintext
Raw Normal View History

NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
#
# For a description of the syntax of this configuration file,
Fix references to the no-longer-existent misc/ directory in comments, README files, and documentation
2015-06-28 22:08:57 +08:00
# see the file kconfig-language.txt in the NuttX tools repository.
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
#
IOBs: Move from driver/iob to a better location in mm/iob
2017-05-09 21:34:59 +08:00
menu "Common I/O Buffer Support"
Clean up some networking configuration menus
2014-06-26 01:01:08 +08:00
IOBs: Move from driver/iob to a better location in mm/iob
2017-05-09 21:34:59 +08:00
config MM_IOB
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
bool "Enable generic I/O buffer support"
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
default n
---help---
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
This setting will build the common I/O buffer (IOB) support
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
library.
IOBs: Move from driver/iob to a better location in mm/iob
2017-05-09 21:34:59 +08:00
if MM_IOB
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
config IOB_NBUFFERS
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
int "Number of pre-allocated I/O buffers"
Make the read ahead buffer unselectable Here is the email loop talk about why it is better to remove the option: https://groups.google.com/forum/#!topic/nuttx/AaNkS7oU6R0 Change-Id: Ib66c037752149ad4b2787ef447f966c77aa12aad Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2020-01-11 11:56:03 +08:00
default 36 if NET_WRITE_BUFFERS
default 24 if !NET_WRITE_BUFFERS && NET_READAHEAD
This commit adds an as-of-yet untested implemented of UDP write buffering. Squashed commit of the following: net/udp: Address most of the issues with UDP write buffering. There is a remaining issue with handling one network going down in a multi-network environment. None of this has been test but it is certainly ready for test. Hence, the feature is marked EXPERIMENTAL. net/udp: Some baby steps toward a corrected write buffering design. net/udp: Remove pesky write buffer macros. Eliminate trailing space at the end of lines. net/udp: A little more UDP write buffering logic. Still at least on big gaping hole in the design. net/udp: Undefined CONFIG_NET_SENDTO_TIMEOUT. net/udp: Crude, naive port of the TCP write buffering logic into UDP. This commit is certainly non-functional and is simply a starting point for the implementatin of UDP write buffering. net/udp: Rename udp/udp_psock_sendto.c udp/udp_psock_sendto_unbuffered.c.
2018-01-23 08:32:02 +08:00
default 8 if !NET_WRITE_BUFFERS && !NET_READAHEAD
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
---help---
Each packet is represented by a series of small I/O buffers in a
chain. This setting determines the number of preallocated I/O
buffers available for packet data.
config IOB_BUFSIZE
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
int "Payload size of one I/O buffer"
Cosmetic plus some configuration default changes
2014-06-13 03:32:07 +08:00
default 196
NET: Add generic I/O buffering logic
2014-06-04 02:41:34 +08:00
---help---
Each packet is represented by a series of small I/O buffers in a
chain. This setting determines the data payload each preallocated
I/O buffer.
modify iob to support header padding and alignment features Signed-off-by: luojun1 <luojun1@xiaomi.com>
2022-08-11 16:54:08 +08:00
config IOB_HEADSIZE
int "Head size of each I/O buffer"
default 0
---help---
This setting determines the reserved size in front of the payload
buffer in each I/O buffer.
config IOB_ALIGNMENT
int "Alignment size of each I/O buffer"
default 1
---help---
The member io_head of all I/O buffers is aligned to the value
specified by this configuration.
IOB: Add queue handling interfaces; improve lists
2014-06-06 23:35:31 +08:00
config IOB_NCHAINS
int "Number of pre-allocated I/O buffer chain heads"
Make the read ahead buffer unselectable Here is the email loop talk about why it is better to remove the option: https://groups.google.com/forum/#!topic/nuttx/AaNkS7oU6R0 Change-Id: Ib66c037752149ad4b2787ef447f966c77aa12aad Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2020-01-11 11:56:03 +08:00
default 0 if !NET_READAHEAD
default IOB_NBUFFERS if NET_READAHEAD
IOB: Add queue handling interfaces; improve lists
2014-06-06 23:35:31 +08:00
---help---
First cut at conversion of write-buffering to use I/O buffer chaings (IOBs)
2014-06-23 01:27:57 +08:00
These tiny nodes are used as "containers" to support queueing of
IOB: Add queue handling interfaces; improve lists
2014-06-06 23:35:31 +08:00
I/O buffer chains. This will limit the number of I/O transactions
First cut at conversion of write-buffering to use I/O buffer chaings (IOBs)
2014-06-23 01:27:57 +08:00
that can be "in-flight" at any give time. The default value of
zero disables this features.
These generic I/O buffer chain containers are not currently used
mm/iob/Kconfig: Fix a typo (other other -> other)
2021-03-15 14:50:25 +08:00
by any logic in NuttX. That is because their other specialized
First cut at conversion of write-buffering to use I/O buffer chaings (IOBs)
2014-06-23 01:27:57 +08:00
I/O buffer chain containers that also carry a payload of usage
specific information.
IOB: Add queue handling interfaces; improve lists
2014-06-06 23:35:31 +08:00
Kconfig: mention that IOB_NCHAINS is not used by TCP anymore
2021-06-30 13:44:23 +08:00
Note: TCP doesn't use this.
Note: UDP and CAN use this.
Add throttle support to the I/O buffer logic
2014-06-25 01:53:19 +08:00
config IOB_THROTTLE
int "I/O buffer throttle value"
This commit adds an as-of-yet untested implemented of UDP write buffering. Squashed commit of the following: net/udp: Address most of the issues with UDP write buffering. There is a remaining issue with handling one network going down in a multi-network environment. None of this has been test but it is certainly ready for test. Hence, the feature is marked EXPERIMENTAL. net/udp: Some baby steps toward a corrected write buffering design. net/udp: Remove pesky write buffer macros. Eliminate trailing space at the end of lines. net/udp: A little more UDP write buffering logic. Still at least on big gaping hole in the design. net/udp: Undefined CONFIG_NET_SENDTO_TIMEOUT. net/udp: Crude, naive port of the TCP write buffering logic into UDP. This commit is certainly non-functional and is simply a starting point for the implementatin of UDP write buffering. net/udp: Rename udp/udp_psock_sendto.c udp/udp_psock_sendto_unbuffered.c.
2018-01-23 08:32:02 +08:00
default 0 if !NET_WRITE_BUFFERS || !NET_READAHEAD
default 8 if NET_WRITE_BUFFERS && NET_READAHEAD
Add throttle support to the I/O buffer logic
2014-06-25 01:53:19 +08:00
---help---
TCP write buffering and read-ahead buffer use the same pool of free
I/O buffers. In order to prevent uncontrolled incoming TCP packets
from hogging all of the available, pre-allocated I/O buffers, a
throttling value is required. This throttle value assures that
I/O buffers will be denied to the read-ahead logic before TCP writes
are halted.
mm/iob: Add an IOB notifier that will send a signal to any registered threads that want to be notified when an IOB has been freed. This is an untested work-in-progress and is intended to be a part of a larger solution to correctly handling network poll operations.
2018-09-09 01:21:18 +08:00
config IOB_NOTIFIER
bool "Support IOB notifications"
default n
sched/wqueue/kwork_notifier.c and several Kconfig files: Notifier should work with either lpwork or hpwork and other minor typo fix.
2019-01-28 01:02:56 +08:00
depends on SCHED_WORKQUEUE
sched/wqueue: Add an option to work queue notifier so that the notification can occur on different work queues.
2018-09-11 21:22:23 +08:00
select WQUEUE_NOTIFIER
mm/iob: Add an IOB notifier that will send a signal to any registered threads that want to be notified when an IOB has been freed. This is an untested work-in-progress and is intended to be a part of a larger solution to correctly handling network poll operations.
2018-09-09 01:21:18 +08:00
---help---
Signals were not a good choice of IPC to implement the poll function for several reasons: In order to handle the asynchrnous poll-related event, a substantial amount of state information is needed. Signals are only capable of passing minimal amounts of data. There are also complexities with performing kernel space signal handlers in kernel space code that is better to avoid. So, instead of signals, the equivalent logic was converted to run via a callback that executes on the high-priority work queue. Squashed commit of the following: Fix up some final compile isses. net/netdev: Convert the network down notification logic to use the new wqueue-based notification factility. net/udp: Convert the UDP readahead notification logic to use the new wqueue-based notification factility. net/tcp: Convert the TCP readahead notification logic to use the new wqueue-based notification factility. mm/iob: Convert the IOB notification logic to use the new wqueue-based notification factility. sched/wqueue: Signals are not good IPCs to support the target poll functionality for several reasons including the amount of data that can be passed with a signal and in the fact that in protected and kernel modes, user threads executing signal handlers in protected, kernel memory is problematic. Instead, convert the same logic to perform the notifications via function callback on the high priority work queue.
2018-09-10 05:01:44 +08:00
Enable building of IOB notifier logic that will execute a worker
function on the high priority work queue when an IOB is available.
This is is a general purpose notifier, but was developed specifically to
support poll() logic where the poll must wait for an IOB to become
available.
mm/iob: Add an IOB notifier that will send a signal to any registered threads that want to be notified when an IOB has been freed. This is an untested work-in-progress and is intended to be a part of a larger solution to correctly handling network poll operations.
2018-09-09 01:21:18 +08:00
mm/iob: Add a divider that can be used to reduce the rate of IOB notifications.
2018-09-13 20:53:58 +08:00
config IOB_NOTIFIER_DIV
int "Notification divider"
default 4
range 1 64
depends on IOB_NOTIFIER
---help---
IOBs may become available at very high rates and the resulting
notification processing can be substantial even if there is nothing
waiting for a free IOB. This divider will reduce that rate of
notification. This must be an even power of two. Supported values
include: 1, 2, 4, 8, 16, 32, 64. The default value of 4 means that
a notification will be sent only when there are a multiple of 4 IOBs
available.
Add throttle support to the I/O buffer logic
2014-06-25 01:53:19 +08:00
config IOB_DEBUG
TCP write buffering: Fix an error in the trim logic
2014-06-23 06:25:26 +08:00
bool "Force I/O buffer debug"
default n
There can be a failure in IOB allocation to some asynchronous behavior caused by the use of sem_post(). Consider this scenario: Task A holds an IOB.  There are no further IOBs.  The value of semcount is zero. Task B calls iob_alloc().  Since there are not IOBs, it calls sem_wait().  The v alue of semcount is now -1. Task A frees the IOB.  iob_free() adds the IOB to the free list and calls sem_post() this makes Task B ready to run and sets semcount to zero NOT 1.  There is one IOB in the free list and semcount is zero.  When Task B wakes up it would increment the sem_count back to the correct value. But an interrupt or another task runs occurs before Task B executes.  The interrupt or other tak takes the IOB off of the free list and decrements the semcount.  But since semcount is then < 0, this causes the assertion because that is an invalid state in the interrupt handler. So I think that the root cause is that there the asynchrony between incrementing the semcount. This change separates the list of IOBs: Currently there is only a free list of IOBs. The problem, I believe, is because of asynchronies due sem_post() post cause the semcount and the list content to become out of sync. This change adds a new 'committed' list: When there is a task waiting for an IOB, it will go into the committed list rather than the free list before the semaphore is posted. On the waiting side, when awakened from the semaphore wait, it will expect to find its IOB in the committed list, rather than free list. In this way, the content of the free list and the value of the semaphore count always remain in sync.
2017-05-17 01:03:35 +08:00
depends on DEBUG_FEATURES && !SYSLOG_BUFFER
TCP write buffering: Fix an error in the trim logic
2014-06-23 06:25:26 +08:00
---help---
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
This option will force debug output from I/O buffer logic. This
is not normally something that would want to do but is convenient
if you are debugging the I/O buffer logic and do not want to get
overloaded with other un-related debug output.
TCP write buffering: Fix an error in the trim logic
2014-06-23 06:25:26 +08:00
Porting Guide: Add description of IOBs.
2017-05-20 22:50:05 +08:00
NOTE that this selection is not available if IOBs are being used
There can be a failure in IOB allocation to some asynchronous behavior caused by the use of sem_post(). Consider this scenario: Task A holds an IOB.  There are no further IOBs.  The value of semcount is zero. Task B calls iob_alloc().  Since there are not IOBs, it calls sem_wait().  The v alue of semcount is now -1. Task A frees the IOB.  iob_free() adds the IOB to the free list and calls sem_post() this makes Task B ready to run and sets semcount to zero NOT 1.  There is one IOB in the free list and semcount is zero.  When Task B wakes up it would increment the sem_count back to the correct value. But an interrupt or another task runs occurs before Task B executes.  The interrupt or other tak takes the IOB off of the free list and decrements the semcount.  But since semcount is then < 0, this causes the assertion because that is an invalid state in the interrupt handler. So I think that the root cause is that there the asynchrony between incrementing the semcount. This change separates the list of IOBs: Currently there is only a free list of IOBs. The problem, I believe, is because of asynchronies due sem_post() post cause the semcount and the list content to become out of sync. This change adds a new 'committed' list: When there is a task waiting for an IOB, it will go into the committed list rather than the free list before the semaphore is posted. On the waiting side, when awakened from the semaphore wait, it will expect to find its IOB in the committed list, rather than free list. In this way, the content of the free list and the value of the semaphore count always remain in sync.
2017-05-17 01:03:35 +08:00
to syslog buffering logic (CONFIG_SYSLOG_BUFFER=y)!
IOBs: Move from driver/iob to a better location in mm/iob
2017-05-09 21:34:59 +08:00
endif # MM_IOB
Move net/iob to drivers/iob so that the I/O buffering feature can be available to other drivers when networking is disabled.
2017-04-21 06:08:49 +08:00
endmenu # Common I/O buffer support