Binutils ld has an annoying misfeature (apparently a regression from a
few years ago) that alignment directives (and alignment specifiers on
symbols) apply only to the runtime addresses and not, apparently, to
the load address region specified with the "AT>" syntax. The net
result is that by default the LMA output ends up too small for the
addresses generated in RAM. See here for some details:
https://sourceware.org/ml/binutils/2013-06/msg00246.htmlhttps://sourceware.org/ml/binutils/2014-01/msg00350.html
The required workaround/fix is that AFAICT any section which can have
inherit a separate VMA vs. LMA from a previous section must specify an
"ALIGN_WITH_INPUT" attribute. Otherwise the sections will get out of
sync and the XIP data will be wrong at runtime.
No, I don't know why this isn't the default behavior.
A further complexity is that this feature only works as advertised
when the section is declared with the "AT> region" syntax after the
block and not "AT(address)" in the header. If you use the header
syntax (with or without ALIGN_WITH_INPUT), ld appears to DOUBLE-apply
padding and the LMA ends up to big. This is almost certainly a
binutils bug, but it's trivial to work around (and the working syntax
is actually cleaner) so we adjust the usage here.
Note finally that this patch includes an effective reversion of commit
d82e9dd9 ("x86: HACK force alignment for _k_task_list section"), which
was an earlier workaround for what seems to be the same issue.
Jira: ZEP-955
Change-Id: I2accd92901cb61fb546658b87d6752c1cd14de3a
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
Linker scripts had not been updated following the addition of
_k_mem_pool sections.
Change-Id: Ic58e893b5296d0f814253e714f8858c272e79913
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
Some modifications to the base linker scripts were not propagated.
Change-Id: I73ab016d861779ad7e633ce8602d2e57845bde85
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
This test checks if it can write at the edges of RAM within the kernel
image. The problem is that this memory is not meant to be trampled on.
With the help of custom linker scripts, place 32-byte buffers at those
edges.
Only linker scripts for QEMU on x86 and Cortex-M3 are provided, to avoid
having to maintain too many of them, in case the reference linker
scripts in the kernel change.
JIRA: ZEP-707
Change-Id: Icd5d680ce2cf064cce083c3d244a196e292bd453
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
This allows running the sanitycheck with:
--tag unified_capable -x KERNEL_TYPE=unified
to run the unified kernel with the tests it is currently known to be
able to run.
Change-Id: Ic145fc6adca162745887672372226fd67447b34a
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
$srctree for the application might not be set to be $ZEPHYR_BASE, use
$ZEPHYR_BASE instead to be more explicit in the build.
Change-Id: Iefa5ff59f246b584949329044f7a6531adc6ed62
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Add more specifications or qualify some to the sanity check test cases
for them to be ran in real hardware:
- kernel types (micro vs nano)
- platforms / arches to exclude / include
- one that is removed (for the PCI sample) as it cannot be ran
without extra information
Change-Id: Id14dc15eb89358c3656d2814ea41bb6fec051278
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
Many test configs are the same, remove complexity and duplication by
using just one kernel config where applicable.
This removes the usage of ARCH which is a remnant from the days where
we had to specify the architecture of the board, the architecture is now
part of Kconfig and determined basded on the board configuration.
This will also make it easy adding new architectures to test cases without
having to add an architecture specific config file when it is actually not
needed, for example now that we will enable micro-kernel support on ARC.
Jira: ZEP-238
Change-Id: I143fa3c4629c58329cfeb0c761c7a896fc1ef63a
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
The old 'config_whitelist' directive in testcase.ini has been removed.
We use the new expr_parser module to parse a 'filter' directive which
is a boolean expression. This gives a great deal more flexibility
in how tests can be filtered.
To keep the tree bisectable, use of config_whitelist in testcase.ini
converted to the new expression language.
Change-Id: I0617319818c5559c0f0569d2fa73d09b681cac51
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
Make the test case routines reside under tests.
Change-Id: Iea59a68e8b537954250d63923a88df267639e716
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Move all kernel testcases to tests/ and change Makefiles
accordingly where applicable.
Change-Id: I130cc3919174e93b7130d55fb101bed1d5d7552d
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Andy Ross