pstore: ramoops: use correct format specifier for size_t variables ...

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: don't try to grab non-existent logbuf ...

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: remove unused goto label ...

Previous refactoring removed jumps to fail_clear. Remove it.

Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: fix build by re-resolving botched merge conflict ...

297ac7288 ("pstore: Only capture log messages") removed the pstore console.
v3 of the eventually merged 75a74b5c ("console: fix out-of-bounds read
in dputc(/dev/*, ...)") reinstated part of this code, which broke the
pstore build because the now removed bufsize member had a user
reintroduced in the patch.

Fix this by reverting the offending patch's changes to the pstore code.
It wasn't applicable to begin with.

Fixes: 75a74b5c ("console: fix out-of-bounds read in dputc(/dev/*, ...)")
Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

console: fix out-of-bounds read in dputc(/dev/*, ...) ...

Trying to output a single character via
	echo -a /dev/serial0-1
currently results in garbage output after the newline, because console.c's
fops_write discards the buffer length and passes the buffer to
(struct cdev)::puts which only handles NUL-terminated strings.

Fix this by amending (struct cdev)::puts with a new nbytes parameter,
which is correctly propagated. All this functions now return at most the
nbytes parameter they were passed in. This fixes __console_puts, which
used to count new lines twice in its return value.

Fixes: b4f55fcf35 ("console: expose consoles in devfs")
Cc: Bastian Krause <bst@pengutronix.de>
Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: remove unused code ...

pstore_info->write() is unused, remove the code implementing it.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Only capture log messages ...

With this pstore only captures barebox log messages printed with pr_*
and dev_*, but no longer anything printed with printf and friends.

When capturing the barebox output with pstore only the log messages are
of interest, but not the ones printed with printf and certainly not
the things typed interactively on the command line. These are logged
currently because we register pstore as a barebox console. Instead, hook
into pr_puts which only outputs the barebox log messages.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: ramoops: allow zapping invalid buffers in read-only mode ...

The FS_PSTORE_RAMOOPS_RO configuration option keeps barebox from zapping
(clearing and fixing header ecc) all ramoops buffers on initialization.
It also stops barebox from zapping invalid buffers. This causes issues
when the console writing code tries to use the uninitialized, invalid
console buffer. Therefore, allow barebox to zap invalid buffers, the
kernel will do so anyway if it finds broken buffers during its
initialization.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: pass ramoops configuration to kernel via device tree ...

Instead of setting ramoops module parameters on the kernel command line,
add a /reserved-memory/ramoops node to the device tree via of_fixup.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Documentation-added-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Replace arguments for write() API ...

Similar to the pstore_info read() callback, there were too many arguments.
This switches to the new struct pstore_record pointer instead. This adds
"reason" and "part" to the record structure as well.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Replace arguments for read() API ...

The argument list for the pstore_read() interface is unwieldy. This changes
passes the new struct pstore_record instead. The erst backend was already
doing something similar internally.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Switch pstore_mkfile to pass record ...

Instead of the long list of arguments, just pass the new record struct.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: add console support ...

Add support for writing console messages to pstore.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Documentation-added-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Extract common arguments into structure ...

The read/mkfile pair pass the same arguments and should be cleared
between calls. Move to a structure and wipe it after every loop.

Signed-off-by: Kees Cook <keescook@chromium.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit 9abdcccc3d5f]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore/ram: Clarify resource reservation labels ...

When ramoops reserved a memory region in the kernel, it had an unhelpful
label of "persistent ram". When reading iomem, it would be repeated many
times, did not hint that it was ramoops in particular, and didn't
clarify very much about what each was used for:

    0x4fdd4000 - 0x4fdf3fff (size 0x00020000) persistent ram
    0x4fdf4000 - 0x4fe13fff (size 0x00020000) persistent ram
    ...
    0x4ff74000 - 0x4ff93fff (size 0x00020000) persistent ram
    0x4ff94000 - 0x4ffb3fff (size 0x00020000) persistent ram
    0x4ffb4000 - 0x4ffd3fff (size 0x00020000) persistent ram

Instead, this adds meaningful labels for how the various regions are
being used:

    0x4fdd4000 - 0x4fdf3fff (size 0x00020000) ramoops:dump(0/12)
    0x4fdf4000 - 0x4fe13fff (size 0x00020000) ramoops:dump(1/12)
    ...
    0x4ff74000 - 0x4ff93fff (size 0x00020000) ramoops:console
    0x4ff94000 - 0x4ffb3fff (size 0x00020000) ramoops:ftrace
    0x4ffb4000 - 0x4ffd3fff (size 0x00020000) ramoops:pmsg

Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit 1227daa43bce]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: improve error report for failed setup ...

When setting ramoops record sizes, sometimes it's not clear which
parameters contributed to the allocation failure. This adds a per-zone
name and expands the failure reports.

Signed-off-by: Kees Cook <keescook@chromium.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit c443a5f3f1f1]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore/ram: Do not use stack VLA for parity workspace ...

Instead of using a stack VLA for the parity workspace, preallocate a
memory region. The preallocation is done to keep from needing to perform
allocations during crash dump writing, etc. This also fixes a missed
release of librs on free.

Signed-off-by: Kees Cook <keescook@chromium.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit f2531f1976d9]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: Make ramoops_init_przs generic for other prz arrays ...

Currently ramoops_init_przs() is hard wired only for panic dump zone
array. In preparation for the ftrace zone array (one zone per-cpu) and pmsg
zone array, make the function more generic to be able to handle this case.

Heavily based on similar work from Joel Fernandes.

Signed-off-by: Kees Cook <keescook@chromium.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit de83209249d6]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

ramoops: use DT reserved-memory bindings ...

Instead of a ramoops-specific node, use a child node of /reserved-memory.
This requires that of_platform_device_create() be explicitly called
for the node, though, since "/reserved-memory" does not have its own
"compatible" property.

Suggested-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Rob Herring <robh@kernel.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit 529182e204db]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore/ram: add Device Tree bindings ...

ramoops is one of the remaining places where ARM vendors still rely on
board-specific shims.  Device Tree lets us replace those shims with
generic code.

These bindings mirror the ramoops module parameters, with two small
differences:

(1) dump_oops becomes an optional "no-dump-oops" property, since ramoops
sets dump_oops=1 by default.

(2) mem_type=1 becomes the more self-explanatory "unbuffered" property.

Signed-off-by: Greg Hackmann <ghackmann@google.com>
[fixed platform_get_drvdata() crash, thanks to Brian Norris]
[switched from u64 to u32 to simplify code, various whitespace fixes]
[use dev_of_node() to gain code-elimination for CONFIG_OF=n]
Signed-off-by: Kees Cook <keescook@chromium.org>
[p.zabel@pengutronix.de: ported to Barebox from Linux commit 35da60941e44]
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

ramoops: probe from device tree if OFTREE is enabled ...

Switch to a device driver probed from device tree if CONFIG_OFTREE is
enabled. Also switch from postcore_initcall to device_initcall, to make
sure that memory banks have been initialized before request_sdram_region
is called.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

fs: devfs: Change .lseek callbacks to return 'int' ...

Returning requested offset from .lseek() callback doesn't really give
us any new information while bringing unnecessary
complications. Change all .lseek() types (both in struct struct
cdev_operations and in struct fs_driver_d) to return 'int' and adjust
the rest of the codebase accordingly.

Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

fs: dentry cache implementation ...

This adds the Linux dentry cache implementation to barebox.

Until now every filesystem driver resolves the full path to a file for
itself. This leads to code duplication and is error prone since
resolving paths is a complicated task. Also it can narrow down the
lookup performance since barebox only knows ASCII paths and has no way
of caching lookups.

With this patch we get the Linux dcache implementation. The path
resolving code from fs/namei.c is nearly taken as-is, minus the RCU
and locking code. Dcaching is made simple as of now: We simply cache
everything and never release any dentries. Although we do reference
counting for inodes and dentries it is effectively not used yet.
We never free anything until a fs is unmounted in which case we free
everything no matter if references are taken or not.

This patch also contains a wrapper in fs/legacy.c to support
filesystems with the old API.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

pstore: lower message priority ...

Having no valid data in pstore after a coldstart is pretty much
expected, so do not print a scary error message, but a debug message
instead.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

remove checks for xzalloc() returning NULL ...

xzalloc() either returns memory or panics, so checking for NULL is useless.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

ramoops: allow to build without OFTREE support ...

Ramoops may be useful even without oftree support, as kernels
booted without a DT may have other means to reserve the
ramoops memory.

Fixes:
In function `ramoops_probe':
undefined reference to `of_add_reserve_entry'

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

Introduce include/linux/mutex.h ...

Move mutex related defines to its original place.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

Introduce include/linux/spinlock.h ...

Move spinlock related definitions to its original place.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

fs: Add pstore filesystem ...

pstore is a persistent storage filesystem used for RAMOOPS. It is used
to store console logs, panics, ftrace and other information in case of a
crash/panic/oops/reboot.

pstore is implemented for barebox as a read-only filesystem at the
moment. It may be extended later on. The idea is to provide a way to
extract essential data from the last running kernel.

Most of the code is copied from the kernel. However this is only a
lightweight implementation without real write support yet.

Signed-off-by: Markus Pargmann <mpa@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>