GitLab

This is the umpteenth incarnation of the McKusick allocator, aimed at
replacing other allocators which have serious issues:

- TLSF is fast but not that good when it comes to memory overhead with
  small sizes (i.e. < 512 bytes) on 64bit.

- heapobj-pshared has decent overhead figures but may be insanely slow
  at deallocating blocks from large, hot heaps.

- the Cobalt core allocator is even worse than heapobj-pshared in
  deallocating blocks, although the system heap should be generally
  small enough not to trigger the most pathological cases in practice,
  hopefully. Nevertheless, the performances are unacceptable.

The original McKusick algorithm implements a quick fit allocation
scheme, based on bucket management of ^2-sized chunks, which
automatically switches to a page-based allocation method for blocks
larger than twice the base page size.

This variant maintains the free page list in AVL trees for fast
lookups of multi-page memory ranges, and pages holding bucketed memory
have a quick allocation bitmap to manage their blocks internally.

The current implementation can replace TLSF for delivering
process-local memory with similar performances but lesser overhead
with small sizes. Most importantly, a kernel variant of HEAPMEM should
replace the Cobalt core allocator. Likewise, heapobj-pshared which is
beyond repair should be replaced with a process-shareable variant as
well, assuming the average size and allocation patterns of real-time
objects are similar in all contexts.

Gilles Chanteperdrix authored Jul 12, 2016 and Philippe Gerum committed Jul 03, 2018

Make the AVL tree usable in shared memory when AVL_SHARED is defined
at build time, switching to offset-based memory references.

Gilles published this code in July 2016 as part of his personal
toolkit for hobby projects aka 'libchutils'.

Some architectures may define a secondary system call range for
internal operations, fix the consistency check not to warn about them.

Jan Kiszka authored May 02, 2018 and Philippe Gerum committed Jul 03, 2018

This stresses the case of fast-path mutex acquire plus pthread_cond_wait
with that mutex, which currently triggers a bug in xnsynch_release.

Along this, drop the HAVE_PTHREAD_MUTEXATTR_SETPROTOCOL conditional -
libboilerplate takes care of that.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored May 02, 2018 and Philippe Gerum committed Jul 03, 2018

We must not try to drop the prio ceiling in the kernel if it wasn't
committed yet. That could easily happen when using a prio-ceiling mutex
to wait on a cond-var.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Norbert Lange authored Apr 25, 2018 and Philippe Gerum committed Jul 03, 2018

Signed-off-by: Norbert Lange <norbert.lange@andritz.com>

Jan Kiszka authored Apr 18, 2018 and Philippe Gerum committed Jul 03, 2018

Otherwise we first add the synch object to the new owner's boost list
before we remove it from the old one's, corrupting the list.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Apr 18, 2018 and Philippe Gerum committed Jul 03, 2018

This triggers a bug in the PP implementation.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Each rtnet device registered with the stack is represented by a node
into /sys/devices/virtual/rtnet/.

Setting the address of the real device probed by the kernel into the
rtnet_device.sysbind field before calling rt_register_rtnetdev(),
causes the "adapter" symlink to be installed in the newly created
attribute directory, pointing at that device.

e.g.

static int probe_handler(struct pci_dev *pdev, const struct pci_device_id *ent)
{
	...
	rtdev->sysbind = &pdev->dev;
	...
	ret = rt_register_rtnetdev(rtdev);
	...
}

$ ls -l /sys/devices/virtual/rtnet/
total 0
drwxr-xr-x    2 root     root             0 Mar 30 12:50 enp1s0
drwxr-xr-x    2 root     root             0 Mar 30 12:50 rtlo

$ ls -l /sys/devices/virtual/rtnet/enp1s0/
total 0
lrwxrwxrwx    1 root     root             0 Mar 30 12:51 adapter -> ../../../pci0000:00/0000:00:13.0/0000:01:00.0
lrwxrwxrwx    1 root     root             0 Mar 30 12:51 subsystem -> ../../../../class/rtnet
-rw-r--r--    1 root     root          4096 Mar 30 12:50 uevent

Henning Schild authored Mar 02, 2018 and Philippe Gerum committed Jul 03, 2018

Moving the definition out of the ifdef solves the following error:

include/xenomai/cobalt/kernel/clock.h:260:16: error:
 ‘struct <anonymous>’ has no member named ‘adjust_time’
include/xenomai/cobalt/kernel/clock.h:263:19: error:
 ‘struct <anonymous>’ has no member named ‘adjust_time’
Signed-off-by: Henning Schild <henning.schild@siemens.com>

CONFIG_X86_X32 used to depend on CONFIG_IA32_EMULATION. Since upstream
commit #9b54050b, this is not the case anymore, and both options can
be enabled separately.

Remove the assumption about such dependency, fixing the build issue.

Jan Kiszka authored Sep 29, 2017 and Philippe Gerum committed Jul 03, 2018

Except for the short phase between thread_init and shadow_map, a thread
is always identifiable via the pid of its Linux mate. Use this shorter
value, which also correlates with what ftrace records anyway, instead of
the pointer or the name. Report the full thread name only in prominent
cases: init, resume and switch.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Sep 29, 2017 and Philippe Gerum committed Jul 03, 2018

All these are synchronous, and the thread context is already recorded by
ftrace.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Sep 29, 2017 and Philippe Gerum committed Jul 03, 2018

Matching numbers against syscall.h, specifically when the call is compat
or x32, is tedious work. Fortunately, ftrace allows up to automate this.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored May 06, 2016 and Philippe Gerum committed Jul 03, 2018

Specifically useful to validate scheduling during PI or PP phases.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Sep 28, 2017 and Philippe Gerum committed Jul 03, 2018

Adding PIDs and the state of the previous task will allow to track
Xenomai task switches in kernelshark (so far via out-of-tree patches,
upstream is planning for the necessary plugin concept).

Moreover, reporting the current priority on context switch helps
debugging unexpected or delayed context switches
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Sep 28, 2017 and Philippe Gerum committed Jul 03, 2018

The code of cobalt_print_sched_params is carried into the format string
in tracefs, and trace-cmd tries to make any sense out of it. While it
can process simply statements, this code is too complex and will prevent
the parsing.

Convert it into a function. That still does not resolve the parsing
issue of trace-cmd, but that can be addressed by a custom plugin which
can then interpret this tracepoint. That wouldn't be possible with the
broken format string.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>

Jan Kiszka authored Sep 28, 2017 and Philippe Gerum committed Jul 03, 2018

__print_symbolic already ensures that unknown policies are printed
numerically.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>