On 06/10/2018 04:21 AM, Amir Goldstein wrote:
On Sun, Jun 10, 2018 at 8:07 AM, Allison Henderson <allison.henderson@xxxxxxxxxx> wrote:Variable sized arrays implemented this way may cause corruptions depending on how different compilers pack the structure. Signed-off-by: Allison Henderson <allison.henderson@xxxxxxxxxx> --- include/parent.h | 1 - io/parent.c | 9 ++++++--- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/include/parent.h b/include/parent.h index f338f96..85cef85 100644 --- a/include/parent.h +++ b/include/parent.h @@ -22,7 +22,6 @@ typedef struct parent { __u64 p_ino; __u32 p_gen; __u16 p_reclen; - char p_name[1];How about char p_name[0]; Is that also risky? keeps the code cleaner IMO. Second best, at least have a macro to access p_name instead of open coding ((char*)parent)+sizeof(struct parent)) all the time. First option is much better IMO. Thanks, Amir.
Oh yes, I think there was some discussion about not using declarations like that in a previous review of the set last year. Initially I had applied the same logic here to sort of prepare these older structs for use in an alternate solution to patch 26, but then later adopted Darricks solution that does not require then. I could probably drop this patch from the set since this code is removed in patch 27. Thx for the catch though!
Allison -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
