Hello Mahesh,
On Tue, 02 Apr 2013 10:43:59 +0530
Mahesh J Salgaonkar <mahesh at linux.vnet.ibm.com> wrote:
> From: Mahesh Salgaonkar <mahesh at linux.vnet.ibm.com>
>
> Recent eppic support in makedumpfile has introduced 'erase_ch' member in
> filter_info structure. This member holds the erase character specified by the
> user to poison the memory location in the dump file that require to be scrubbed
> out. But, when the memory range to be scrubbed out is spanned across multiple
> pages, only the memory that belongs to first page is poisoned with the
> character set in 'erase_ch' member. The rest of the memory gets poisoned with
> 'null' character. This is because, the 'erase_ch' member is not copied over
> while splitting filter info in split_filter_info() function.
>
> Signed-off-by: Mahesh Salgaonkar <mahesh at linux.vnet.ibm.com>
> Signed-off-by: Suzuki Poulose <suzuki at in.ibm.com>
> ---
> erase_info.c | 11 +++++++----
> 1 file changed, 7 insertions(+), 4 deletions(-)
>
> diff --git a/erase_info.c b/erase_info.c
> index 61fe637..a1b08e0 100644
> --- a/erase_info.c
> +++ b/erase_info.c
> @@ -1921,12 +1921,15 @@ split_filter_info(struct filter_info *prev, unsigned long long next_paddr,
> ERRMSG("Can't allocate memory to split filter info\n");
> return;
> }
> - new->nullify = prev->nullify;
> - new->erase_info_idx = prev->erase_info_idx;
> - new->size_idx = prev->size_idx;
> +
> + /*
> + * copy over existing data from prev node and only update fields
> + * that differ. This approach will take care of copying over of any
> + * future member addition to filter_info structure.
> + */
> + *new = *prev;
> new->paddr = next_paddr;
> new->size = size;
> - new->next = prev->next;
> prev->next = new;
> }
Sounds good!
I'll merge this patch into v1.5.4.
Thanks
Atsushi Kumagai
