[PATCH] env: Fix bgenv_create_new() behavior on same revisions

[Christian Storm]

From: Christian Storm <christi...@siemens.com>

If both two environments have the same revision, then
bgenv_open_latest() and bgenv_open_oldest() return the
exact same BGENV which then gets reset, losing the
kernel, kernelfile, and watchdog timeout information.

The control flow-following memcpy() in ebg_env_create_new()
at env/env_api.c:86, intended to preserve this information,
is futile.

Hence, only reset the env if latest and oldest differ.

Signed-off-by: Christian Storm <christi...@siemens.com>
---
env/env_api_fat.c | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/env/env_api_fat.c b/env/env_api_fat.c
index d588e7f..503dedf 100644
--- a/env/env_api_fat.c
+++ b/env/env_api_fat.c
@@ -437,20 +437,22 @@ BGENV *bgenv_create_new(void)

int new_rev = env_latest->data->revision + 1;

- bgenv_close(env_latest);
-
env_new = bgenv_open_oldest();
if (!env_new) {
+ bgenv_close(env_latest);
goto create_new_io_error;
}

- /* zero fields */
- memset(env_new->data, 0, sizeof(BG_ENVDATA));
+ if (env_latest->data != env_new->data) {
+ /* zero fields */
+ memset(env_new->data, 0, sizeof(BG_ENVDATA));
+ /* set default watchdog timeout */
+ env_new->data->watchdog_timeout_sec = DEFAULT_TIMEOUT_SEC;
+ }
+ bgenv_close(env_latest);
/* update revision field and testing mode */
env_new->data->revision = new_rev;
env_new->data->in_progress = 1;
- /* set default watchdog timeout */
- env_new->data->watchdog_timeout_sec = DEFAULT_TIMEOUT_SEC;

return env_new;

--
2.40.1

[Jan Kiszka]

Thanks, applied.

Jan

--
Siemens AG, Technology
Competence Center Embedded Linux