This makes discovery of the Module.symvers happen at model build time,
instead of at build tree creation/configuration.

A find_module_symvers script is included to find the Module.symvers
file, and it's used in the RCG Makefile to find module symvers files
at build time.

This reverts the top-level configure script changes from
8c01de98, restoring the configure
script and dropping the autoconf-generated configure step from the
build setup.

remove unnecessary sudo call from RCG install script

See merge request cds/advligorts!14

Fix feCodeGen.pl for timing slaves

Closes #28

See merge request cds/advligorts!39

Enable kernel module versioning - Rev2

Closes #14

See merge request cds/advligorts!40

post_build_script white space clean up

See merge request cds/advligorts!43

remove old run script for FE EDCU

See merge request cds/advligorts!44

Remove RCG 2.4 upgrade scripts

Closes #19 and #31

See merge request cds/advligorts!38

Close #31.  This removes scripts created to migrate ws,front-end to RCG 2.4.
They are no longer applicable as we go to more package, cmake based solutions

Closes #19 - Remove the old CSS Studio files used for creating user-interface (BOY) files
Remove plugins directory with pv and reader.nds. Remove jars directory with old xml
Remove commented-out sections of install scripts that used to make custom OPI files

Close #28 - Add in code so feCodeGen.pl sets number of ADCs correctly
for VirtualIOP machines - if ADCMASTER (IOP) - set = 1, otherwise set =0

Jameson Rollins authored 5 years ago and Jameson Rollins committed 5 years ago

no functional change, but script was broken due to mixture of
tabs and spaces in indentation.

rather than one maintained by the now-deprecated ietf-leap-seconds package.

CDS Controls authored 5 years ago and Jonathan Hanks committed 5 years ago

Commiting the work Patrick Thomas had done for moving the ini file crc to a background thread on the edcu

Closes #14 - Restore support for kernel module versioning
patches/kernel-4.19-cs/config-4.19.67-cs - change to MODVERSIONS=y
patches/kernel-4.19-cs/README - Update note on config file
-- changes for mbuf, gpstime symbol versions --
Update RCG autotools - remove configure, move configure.in to configure.ac
bootstrap - add to generate configure from configure.ac
configure.ac - rename from configure.in, add tests to find path to mbuf, gpstime
Module.symvers files - either /var/lib/dkms/$MODNAME/$MODVERS/$KVERS/$ARCH/
for DKMS package or /var/cache/$MODNAME from local build
src/drv/ExtraSymbols.symvers - Remove mbuf, gpstime symbols
src/drv/gpstime/Makefile - Add caching of Module.symvers in /var/cache/gpstime on install
src/drv/mbuf/Makefile - Add caching of Module.symvers in /var/cache/mbuf on install
src/epics/util/Makefile.in  - Add mbuf,gpstime Module.symvers paths
src/epics/util/feCodeGen.pl - Add KBUILD_EXTRA_SYMBOLS statements to front-end make file
-- changes to ...

Closes #15 Remove kernel patch changing cpu_up to EXPORT_SYMBOL
Update README in patches/linux-4.19-cs for change
Remove symbols from patched kernel in src/drv/ExtraSymbols.symvers
Add parameter in GNUmakefile to build will reference Module.symvers from kernel

Removing some unused diagnostic messages that could overflow under large dcu counts and corrupt the stack

The code had been inserting a dcuid '0' into the output headers, make it just skip the model completely in the output.
Also removed a duplicate function to parse a comma delimited list of integers.

Disable CONFIG_DRM_AST flag so kernel uses 'nomodeset' on AST VGA chips
which prevents interrupts causing long IOP cycles on newer front-ends
(and removes need for nomodeset as kernel boot parameter)
Update README in patches/linux-4.19-cs for these changes

Update the test code to allow specifying specific dcus.
Update local_dc to fail if it cannot determine the model rate.

* Increases the DAQ_TRANSIT_MAX_DCU to 256
* Updates the tests of daqd/nds to run with more dcus and sample more channels
* Fixes a spurious failure in the edcu test to keep tests passing

This fixes a bug in which arbitrary waveform excitations failed, with awgtpman reporting the data was too far in the future

Allow spurious failures.  Just require that we exceed the number of successful connections that used to be in a 32 bit bit mask.

The FRAMECPP_VERSION_NUMBER is not properly defined in v2.7.0.  If it is empty, assume version 2.7.0.

Allow spurious failures.  Just require that we exceed the number of successful connections that used to be in a 32 bit bit mask.

This addresses the issue by expanding the size of the bitmask.
* Replace the 'int' bitmask with a generic structure.  This allows the
type system to help us find all the places it is used.
* Replace direct bitmask manipulation with member functions to increase
local readability.
* Allow the new bitmask structure to have an increased capacity that can
be set (at compile time).
* Add unit tests to new bitmask code.
* Add test code to look at nds1 load.