At present panic() is in the vsprintf.h header file. That does not seem
like an obvious choice for hang(), even though it relates to panic(). So
let's put hang() in its own header.
Signed-off-by: Simon Glass <sjg@chromium.org>
[trini: Migrate a few more files]
Signed-off-by: Tom Rini <trini@konsulko.com>
These don't need to be in common.h so move them out into a new header.
Also add some missing comments.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
At present bootstage relocation assumes that it is possible to point back
to memory available before relocation, so it does not relocate the
strings. However this is not the case on some platforms, such as x86 which
uses the cache as RAM and loses access to this when the cache is enabled.
Move the relocation step to before U-Boot relocates, expand the allocated
region to include space for the strings and relocate the strings at the
same time as the bootstage records.
This ensures that bootstage data can remain accessible from TPL through
SPL to U-Boot before/after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present there is a single shared address for bootstage data in both
TPL and SPL. If SPL unstashs TPL bootstage info and then stashes it again
to pass it to U-Boot, the new stash overwrites the strings of the old
stash.
Fix this by duplicating the strings into the malloc() region. This should
be a small code. Fix the header-file order at the same time.
This problem doesn't happen at the next stage (SPL->U-Boot) since U-Boot
relocates the boostage data.
Signed-off-by: Simon Glass <sjg@chromium.org>
The current code searches for empty records but these not existing with
bootstage now. This used to be needed when bootstage records were stored
in a spare array.
Drop the unnecessary code and fix a code-style nit at the same time.
Signed-off-by: Simon Glass <sjg@chromium.org>
When stashing bootstage info, store the next ID so that it can be used
when the stash is restored. This avoids the ID starting at zero and
potentially overwriting existing entries.
Signed-off-by: Simon Glass <sjg@chromium.org>
It is possible for this to happen if something goes wrong very early in
the init sequence. Add a check for this.
Signed-off-by: Simon Glass <sjg@chromium.org>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
Thomas reported U-Boot failed to build host tools if libfdt-devel
package is installed because tools include libfdt headers from
/usr/include/ instead of using internal ones.
This commit moves the header code:
include/libfdt.h -> include/linux/libfdt.h
include/libfdt_env.h -> include/linux/libfdt_env.h
and replaces include directives:
#include <libfdt.h> -> #include <linux/libfdt.h>
#include <libfdt_env.h> -> #include <linux/libfdt_env.h>
Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
With SPL we often have limited memory and do not need very many bootstage
records. Add a separate Kconfig option for SPL.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The CONFIG_BOOTSTAGE_USER_COUNT option is no-longer needed since we can now
support any number of user IDs. Also BOOTSTAGE_ID_COUNT is not needed now.
Drop these unused options.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
There are a few places that should use const *, such as
bootstage_unstash(). Update these to make it clearer when parameters are
changed.
Signed-off-by: Simon Glass <sjg@chromium.org>
We should return a proper error number instead of just -1. This helps the
caller to determine what when wrong. Update a few functions to fix this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards cannot access pre-relocation data after relocation. Reserve
space for this and copy it during preparation for relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
We can now use the record count to determine whether a record is valid or
not. Drop the test for a zero time.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present bootstage has a large array with all possible bootstage IDs
recorded. It adds times to the array element indexed by the ID. This is
inefficient because many IDs are not used during boot. We can save space
by only recording those IDs which actually have timestamps.
Update the array to use a record count, which increments with each
addition of a new timestamp. This takes longer to record a time, since it
may involve an array search. Such a search may be particularly expensive
before relocation when the CPU is running slowly or the cache is off. But
at that stage there should be very few records.
Signed-off-by: Simon Glass <sjg@chromium.org>
There are several code style and comment nits. Fix them and also remove
the comment about passing bootstage to the kernel being TBD. This is
already supported.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present bootstage uses the data section of the image to store its
information. There are a few problems with this:
- It does not work on all boards (e.g. those which run from flash before
relocation)
- Allocated strings still point back to the pre-relocation data after
relocation
Now that U-Boot has a pre-relocation malloc() we can use this instead,
with a pointer to the data in global_data. Update bootstage to do this and
set up an init routine to allocate the memory.
Now that we have a real init function, we can drop the fake 'reset' record
and add a normal one instead.
Note that part of the problem with allocated strings remains. They are
reallocated but this will only work where pre-relocation memory is
accessible after relocation.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present we provide a default version of this function for use by
bootstage. However it uses the system timer and therefore likely requires
driver model. This makes it impossible to time driver-model init.
Drop the function and require boards to provide their own. Add a sandbox
version also. There is a default implememtation in lib/time.c for boards
which use CONFIG_SYS_TIMER_COUNTER.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a function which allows a (file, function, line number) to be marked
in bootstage.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Che-Liang Chiou <clchiou@chromium.org>
Any pointers to name strings that were passed to bootstage_mark_name()
pre-relocation should be copied post-relocation so that they don't get
trashed as the original location of U-Boot is re-used for other
purposes.
This change introduces a new API call that should be called from
board_init_r() after malloc has been initted on any board that uses
bootstage.
Signed-off-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
It is useful to be able to write the bootstage information to memory for
use by a later utility, or the Linux kernel. Provide a function to do
this as well as a function to read bootstage information back and incorporate
it into the current table.
This also makes it possible for U-Boot to chain to another U-Boot and pass
on its bootstage information.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add an option, CONFIG_BOOTSTAGE_FDT to pass boot timings to the kernel
in the device tree, if available. To use this, you must have
CONFIG_OF_LIBFDT defined.
Signed-off-by: Simon Glass <sjg@chromium.org>
Sometimes we want to add up the amount of time spent in a particular
activity when it is happening in a number of discrete chunks.
Add bootstage_start() to mark the start of an acitivity and
bootstage_accum() to accumulate the time since the last start. Calling
these function in pairs results in the accumulated time being collected.
Signed-off-by: Simon Glass <sjg@chromium.org>
This function is not static, but not exported either. Add a prototype
in the header file and move the required enum to the header also.
Signed-off-by: Simon Glass <sjg@chromium.org>
The standalone example does not have get_timer() defined, so we cannot
rely on it being available.
Move the timer function into boootstage.c to avoid this problem.
This corrects a build breakage for the standalone example on some boards.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Matthias Fuchs <matthias.fuchs@esd.eu>
This defines the basics of a new boot time measurement feature. This allows
logging of very accurate time measurements as the boot proceeds, by using
an available microsecond counter.
To enable the feature, define CONFIG_BOOTSTAGE in your board config file.
Also available is CONFIG_BOOTSTAGE_REPORT which will cause a report to be
printed just before handing off to the OS.
Most IDs are not named at this stage. For that I would first like to
renumber them all.
Timer summary in microseconds:
Mark Elapsed Stage
0 0 reset
205,000 205,000 board_init_f
6,053,000 5,848,000 bootm_start
6,053,000 0 id=1
6,058,000 5,000 id=101
6,058,000 0 id=100
6,061,000 3,000 id=103
6,064,000 3,000 id=104
6,093,000 29,000 id=107
6,093,000 0 id=106
6,093,000 0 id=105
6,093,000 0 id=108
7,089,000 996,000 id=7
7,089,000 0 id=15
7,089,000 0 id=8
7,097,000 8,000 start_kernel
Signed-off-by: Simon Glass <sjg@chromium.org>
