AG patches chunk A

Thu Feb 2 14:54:46 UTC 2012

On 02/02/2012 08:39 AM, Daniel Hellstrom wrote:
> Hello,
>
> I summarize the first chunk of patches, PATCH means I will send them on the rtems-devel list, PR means I will file a bugzilla PR. Feel free to comment.. and apply :)
>
I am putting my first thought on whether or not a PR is required.
Correct me if I am wrong.
> Patches are listed in reverse apply order, thus IO_MANAGER patch comes last.
>
> Thanks,
> Daniel
>
>
> + SPARC specific enhancements
>       c1afe98 LEON3: fix compiler warning in SHM driver
>
This sounds like an easy one to review. Probably completely
benign and applies to 4.10 branch. Bugzilla.

> + SPARC specific bugs
>       3e2aa5e LEON3BSP MP: may wake one more CPU than expected
Head only.
>       fe9dc9e SPARC BSPs: workspace must be aligned on a 8-byte boundary
Bugzilla.
>       124e68d LEON3: CPU index intialization moved to bspstart.c
I thought head but I read your description. This sounds like
a bug fix to include in 4.10.
>       c7d92b6 LEON3: change d-cache snoop detect implementation
Your description below makes this sound like it is a bug which
needs to be a Bugzilla PR so it can be in a 4.10 release.
>       LEON specific because GRETH driver only used by LEON:
>       7a92dba GRETH: Moved print to remove potential deadlock
>       6c66b13 GRETH: performance improvements and one bugfix
>
Treat as one PR in Bugzilla.
> + General bugs, (but not in any release, so PR not filed)
>       839ef7b MP06 TEST: avoid program failure when slow UART on CPU core 0
>
How is this not in a release?
> + General enhancements
>       fe2ee40 bootstrap: handy --root switch for bootstrapping subdirs
>       11ae399 IO_MANAGER: early dynamic driver registration
>
>
Definitely through devel.
> PATCH commit 8713dbeead0fb5ce880793ef391916cc422efce5
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Thu Nov 24 16:54:28 2011 +0100
>
>       IO_MANAGER: early dynamic driver registration
>
>       Adds the possibility to register drivers before the IO Manager has
>       completed the initialization. Sometimes the BSP may want to register
>       a driver dynamically early in the boot process.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PATCH commit b2a54a1af400bf731855fe26c8c4550a7cc0697b
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Nov 25 13:58:07 2011 +0100
>
>       bootstrap: handy --root switch for bootstrapping subdirs
>
>       Use the "--root DIR" to process only a subdir, for example
>        $ ./bootstrap --root c/src/lib/libbsp/sparc
>       will only generate Makefiles for all SPARC BSPs, saves some
>       time when working with a specific BSP.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR commit 6c66b134e18dc24b3fe8b2d9a9a6954c6bc6840e
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Dec 2 10:47:58 2011 +0100
>
>       GRETH: performance improvements and one bugfix
>
>       GRETH driver updated, 10-15% performance improvements for GBIT MAC,
>       unnecessary RX interrupts not taken which under heavy load saves approx.
>       1500 interrupts/s, one task removed saving about 5kb memory and 1 bug
>       solved.
>
>       BUG: RX interrupt was enabled before the RX-daemon was created which could
>       result in a faulty call to rtems_event_send.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR commit 7a92dba76fa113ab9fba1ddc90a2578ea8ff8d0a
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Dec 2 10:45:10 2011 +0100
>
>       GRETH: Moved print to remove potential deadlock
>
>       Deadlock may arise when the EDCL bug link is used to tunnel
>       console output over Ethernet, when Ethernet is down one should
>       avoid using console (only during debugging of LEON targets)
>
>       Author: Marko Isomaki<marko at gaisler.com>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR[DEP-1] commit c7d92b6562014f85a29bd94ceb2c3292f36dfb32
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Nov 25 14:02:00 2011 +0100
>
>       LEON3: change d-cache snoop detect implementation
>
>       The previous code only checked if d-cache snooping was implemented,
>       however snooping may be available but not enabled which may lead
>       to driver bugs.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR commit 124e68d114f1ae6489ee86167357e6b6b16503e5
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Nov 25 14:15:16 2011 +0100
>
>       LEON3: CPU index intialization moved to bspstart.c
>
>       All LEON3/4 systems have a CPU-id, if on a single-CPU system the
>       ID is always zero. On a multicore system it ranges from 0 to 15.
>
>       The CPU index should always by updated even in a non-MP RTEMS OS
>       since the CPU running RTEMS may not always be CPU0. For example
>       when RTEMS runs on CPU1 and Linux on CPU0 in a mixed ASMP system.
>
>       The old code executed within the IRQ controller initialization code
>       makes no sense since the ASR register is a CPU register, it has
>       nothing to do with AMBA initialization either.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PATCH commit 839ef7ba191dc65ecbb777bbd108758025259d34
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Wed Nov 16 16:56:46 2011 +0100
>
>       MP06 TEST: avoid program failure when slow UART on CPU core 0
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR commit fe9dc9e1a55c1e71980eaa290f193cdae79141ca
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Nov 25 10:54:28 2011 +0100
>
>       SPARC BSPs: workspace must be aligned on a 8-byte boundary
>
>       The linker-script already declares a 8-byte aligned symbol (_end)
>       which can be used instead of "variable aligned" symbol end.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PR commit 3e2aa5ed2415ff02dac7545930f0a07c546dac77
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Wed Nov 9 16:09:55 2011 +0100
>
>       LEON3BSP MP: may wake one more CPU than expected
>
>       The SHM code always wakes one CPU more that configured, however
>       this has never been a problem since RTEMS will be running on all CPUs
>       or only two cores were available.
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> PATCH commit c1afe981bf63c79f695b6ff826c5109a9f7f1dd7
> Author: Daniel Hellstrom<daniel at gaisler.com>
> Date:   Fri Nov 25 11:33:11 2011 +0100
>
>       LEON3: fix compiler warning in SHM driver
>
>       Signed-off-by: Daniel Hellstrom<daniel at gaisler.com>
>
> _______________________________________________
> rtems-devel mailing list
> rtems-devel at rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-devel


-- 
Joel Sherrill, Ph.D.             Director of Research&   Development
joel.sherrill at OARcorp.com        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
     Support Available             (256) 722-9985