From: Vasile-Laurentiu Stanimir <[email protected]>
While kmsg_bytes can be set for pstore via mount, if a crash occurs
before the mount only partial console log will be stored as kmsg_bytes
defaults to a potentially different hardcoded value in the kernel
(PSTORE_DEFAULT_KMSG_BYTES). This makes it impossible to analyze valuable
post-mortem data especially on the embedded development or in the process
of bringing up new boards. Change this value to be a Kconfig option
with the default of old PSTORE_DEFAULT_KMSG_BYTES
Signed-off-by: Vasile-Laurentiu Stanimir <[email protected]>
Signed-off-by: Kees Cook <[email protected]>
---
This was sent off-list, so I'm resending it to lkml (with the commit log
cleaned up sligthly) before I push it into for-next/pstore.
---
fs/pstore/Kconfig | 8 ++++++++
fs/pstore/inode.c | 2 +-
fs/pstore/internal.h | 1 -
fs/pstore/platform.c | 2 +-
4 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/fs/pstore/Kconfig b/fs/pstore/Kconfig
index e16a49ebfe54..8adabde685f1 100644
--- a/fs/pstore/Kconfig
+++ b/fs/pstore/Kconfig
@@ -14,6 +14,14 @@ config PSTORE
If you don't have a platform persistent store driver,
say N.
+config PSTORE_DEFAULT_KMSG_BYTES
+ int "Default kernel log storage space" if EXPERT
+ depends on PSTORE
+ default "10240"
+ help
+ Defines default size of pstore kernel log storage.
+ Can be enlarged if needed, not recommended to shrink it.
+
config PSTORE_DEFLATE_COMPRESS
tristate "DEFLATE (ZLIB) compression"
default y
diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c
index c331efe8de95..93a217e4f563 100644
--- a/fs/pstore/inode.c
+++ b/fs/pstore/inode.c
@@ -266,7 +266,7 @@ static void parse_options(char *options)
*/
static int pstore_show_options(struct seq_file *m, struct dentry *root)
{
- if (kmsg_bytes != PSTORE_DEFAULT_KMSG_BYTES)
+ if (kmsg_bytes != CONFIG_PSTORE_DEFAULT_KMSG_BYTES)
seq_printf(m, ",kmsg_bytes=%lu", kmsg_bytes);
return 0;
}
diff --git a/fs/pstore/internal.h b/fs/pstore/internal.h
index 7fb219042f13..801d6c0b170c 100644
--- a/fs/pstore/internal.h
+++ b/fs/pstore/internal.h
@@ -6,7 +6,6 @@
#include <linux/time.h>
#include <linux/pstore.h>
-#define PSTORE_DEFAULT_KMSG_BYTES 10240
extern unsigned long kmsg_bytes;
#ifdef CONFIG_PSTORE_FTRACE
diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c
index 36714df37d5d..32f64abc277c 100644
--- a/fs/pstore/platform.c
+++ b/fs/pstore/platform.c
@@ -101,7 +101,7 @@ static char *big_oops_buf;
static size_t big_oops_buf_sz;
/* How much of the console log to snapshot */
-unsigned long kmsg_bytes = PSTORE_DEFAULT_KMSG_BYTES;
+unsigned long kmsg_bytes = CONFIG_PSTORE_DEFAULT_KMSG_BYTES;
void pstore_set_kmsg_bytes(int bytes)
{
--
2.25.1
On Tue, Dec 01, 2020 at 12:12:34PM -0800, Kees Cook wrote:
> ---
> This was sent off-list, so I'm resending it to lkml (with the commit log
> cleaned up sligthly) before I push it into for-next/pstore.
Also, nice:
Writing /tmp/20201201_keescook_pstore_move_kmsg_bytes_default_into_kconfig.mbx
✔ [PATCH] pstore: Move kmsg_bytes default into Kconfig
---
✔ Attestation-by: Kees Cook <[email protected]> (pgp: 8972F4DFDC6DC026)
Did you use the sendemail-validate hook for this?
:)
On Wed, Dec 02, 2020 at 04:42:57PM -0500, Konstantin Ryabitsev wrote:
> On Tue, Dec 01, 2020 at 12:12:34PM -0800, Kees Cook wrote:
> > ---
> > This was sent off-list, so I'm resending it to lkml (with the commit log
> > cleaned up sligthly) before I push it into for-next/pstore.
>
> Also, nice:
>
> Writing /tmp/20201201_keescook_pstore_move_kmsg_bytes_default_into_kconfig.mbx
> ✔ [PATCH] pstore: Move kmsg_bytes default into Kconfig
> ---
> ✔ Attestation-by: Kees Cook <[email protected]> (pgp: 8972F4DFDC6DC026)
Yay! :)
> Did you use the sendemail-validate hook for this?
In my scripts right now, I'm doing this before "git send-email":
# Construct header-based attestations
b4 attest outgoing/*
I haven't yet converted to using the git hook.
--
Kees Cook
On Wed, Dec 02, 2020 at 02:30:08PM -0800, Kees Cook wrote:
> > Did you use the sendemail-validate hook for this?
>
> In my scripts right now, I'm doing this before "git send-email":
>
> # Construct header-based attestations
> b4 attest outgoing/*
>
> I haven't yet converted to using the git hook.
You don't have to if you're already using wrapper scripts. It's more
useful for folks who call git-send-email directly and don't want to
remember to call "b4 attest" first.
Best regards,
-K