Hi
marcandre.lureau@redhat.com writes:
> From: Marc-André Lureau <marcandre.lureau@redhat.com>
>
> Since g_input_stream_read_all() may return less than requested when the
> stream is malformed, we should treat this condition as a runtime user
> error (g_return are for programming errors).
>
> Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
> ---
> backends/dbus-vmstate.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/backends/dbus-vmstate.c b/backends/dbus-vmstate.c
> index bd050e8e9c..616d291cfb 100644
> --- a/backends/dbus-vmstate.c
> +++ b/backends/dbus-vmstate.c
> @@ -229,7 +229,10 @@ static int dbus_vmstate_post_load(void *opaque, int version_id)
> &bytes_read, NULL, &err)) {
> goto error;
> }
> - g_return_val_if_fail(bytes_read == len, -1);
> + if (bytes_read != len) {
> + error_report("%s: Failed to read proxy Id", __func__);
Error messages containing function names are code smell. It's
consustent with nearby errors, i.e. this patch is not to blame.
> + return -1;
> + }
> id[len] = 0;
>
> trace_dbus_vmstate_loading(id);
Reviewed-by: Markus Armbruster <armbru@redhat.com>
For the record, Markus sent a similar patch later "backends/dbus-vmstate: Fix short read error handling", which has been applied.
--
Marc-André Lureau