kernel.git/fs/pstore/ram.c, branch linux-4.8.y

pstore/ramoops: fixup driver removal

2016-10-28T07:45:20Z

commit 4407de74df18ed405cc5998990004c813ccfdbde upstream. A basic rmmod ramoops segfaults. Let's see why. Since commit 34f0ec82e0a9 ("pstore: Correct the max_dump_cnt clearing of ramoops") sets ->max_dump_cnt to zero before looping over ->przs but we didn't use it before that either. And since commit ee1d267423a1 ("pstore: add pstore unregister") we free that memory on rmmod. But even then, we looped until a NULL pointer or ERR. I don't see where it is ensured that the last member is NULL. Let's try this instead: simply error recovery and free. Clean up in error case where resources were allocated. And then, in the free path, rely on ->max_dump_cnt in the free path. Cc: Anton Vorontsov Cc: Colin Cross Cc: Kees Cook Cc: Tony Luck Cc: Namhyung Kim Acked-by: Namhyung Kim Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman

ramoops: use persistent_ram_free() instead of kfree() for freeing prz

2016-08-05T18:21:46Z

persistent_ram_zone(=prz) structures are allocated by persistent_ram_new(), which includes vmap() or ioremap(). But they are currently freed by kfree(). This uses persistent_ram_free() for correct this asymmetry usage. Signed-off-by: Hiraku Toyooka Signed-off-by: Nobuhiro Iwamatsu Cc: Mark Salyzyn Cc: Seiji Aguchi Signed-off-by: Kees Cook

ramoops: use DT reserved-memory bindings

2016-08-05T18:21:36Z

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 Signed-off-by: Kees Cook Acked-by: Rob Herring

pstore/ram: add Device Tree bindings

2016-06-14T18:34:39Z

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 [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

pstore: add lzo/lz4 compression support

2016-06-02T17:59:31Z

Like zlib compression in pstore, this patch added lzo and lz4 compression support so that users can have more options and better compression ratio. The original code treats the compressed data together with the uncompressed ECC correction notice by using zlib decompress. The ECC correction notice is missing in the decompression process. The treatment also makes lzo and lz4 not working. So I treat them separately by using pstore_decompress() to treat the compressed data, and memcpy() to treat the uncompressed ECC correction notice. Signed-off-by: Geliang Tang Signed-off-by: Kees Cook

pstore: Add support for 64 Bit address space

2016-03-10T17:43:36Z

Some architectures have their reserved RAM in 64 Bit address space. Therefore convert mem_address module parameter to ullong. Signed-off-by: Wladislav Wiebe Acked-by: Kees Cook Signed-off-by: Tony Luck

pstore: add pstore unregister

2015-10-22T15:59:18Z

pstore doesn't support unregistering yet. It was marked as TODO. This patch adds some code to fix it: 1) Add functions to unregister kmsg/console/ftrace/pmsg. 2) Add a function to free compression buffer. 3) Unmap the memory and free it. 4) Add a function to unregister pstore filesystem. Signed-off-by: Geliang Tang Acked-by: Kees Cook [Removed __exit annotation from ramoops_remove(). Reported by Arnd Bergmann] Signed-off-by: Tony Luck

ramoops: make it possible to change mem_type param.

2015-05-21T16:52:29Z

If we set ramoops.mem_type=1 in command line, the current code can not change mem_type to 1, because it is assigned to 0 in function ramoops_register_dummy. This patch make it possible to change mem_type parameter in command line. Signed-off-by: Wang Long Acked-by: Tony Lindgren Signed-off-by: Tony Luck

pstore/ram: verify ramoops header before saving record

2015-05-21T16:50:34Z

On some devices the persistent memory contains junk after a cold boot, and /dev/pstore/dmesg-ramoops-* are created with random data which is not the result of a kernel crash. This patch adds a ramoops header check and skips any persistent_ram_zone that does not have a valid header. Signed-off-by: Ben Zhang Signed-off-by: Tony Luck

fs/pstore: Optimization function ramoops_init_przs

2015-05-21T16:47:22Z

The value of cxt->record_size does not change in the loop, so this patch optimize the assign statement by dropping sz entirely and using cxt->record_size in its place. Signed-off-by: Wang Long Acked-by: Kees Cook Signed-off-by: Tony Luck