Hi Vinod,
Thanks for the review,
On 08/07/2018 11:45 AM, Vinod wrote:
Hi Sibi,
On 27-07-18, 20:49, Sibi Sankar wrote:
Introduce custom dump function per remoteproc segment. It is responsible
for filling the device memory segment associated with coredump
Signed-off-by: Sibi Sankar <si...@codeaurora.org>
---
drivers/remoteproc/remoteproc_core.c | 15 ++++++++++-----
include/linux/remoteproc.h | 3 +++
2 files changed, 13 insertions(+), 5 deletions(-)
diff --git a/drivers/remoteproc/remoteproc_core.c
b/drivers/remoteproc/remoteproc_core.c
index 283b258f5e0f..ec56cd822b26 100644
--- a/drivers/remoteproc/remoteproc_core.c
+++ b/drivers/remoteproc/remoteproc_core.c
@@ -1183,13 +1183,18 @@ static void rproc_coredump(struct rproc *rproc)
phdr->p_align = 0;
ptr = rproc_da_to_va(rproc, segment->da, segment->size);
- if (!ptr) {
- dev_err(&rproc->dev,
+
+ if (segment->dump) {
+ segment->dump(rproc, ptr, segment->size, data + offset);
Am not sure I follow, you are calling this w/o checking if ptr is valid,
so you maybe passing null to segment->dump() ?
the rationale behind passing ptr directly to dump_fn is that it will
help in tracking the segments being core dumped (q6v5_pil in particular
requires to unlock mba before dumping and cleanup after all the segments
are dumped which is currently decided based on a mask that is
maintained). It also allows the remoteproc driver to fill the memory as
needed (instead of the default 0xff). This is applicable to drivers that
implement dump_fn, for others the default behavior is maintained.
+ } else {
+ if (!ptr) {
+ dev_err(&rproc->dev,
"invalid coredump segment (%pad, %zu)\n",
&segment->da, segment->size);
- memset(data + offset, 0xff, segment->size);
- } else {
- memcpy(data + offset, ptr, segment->size);
+ memset(data + offset, 0xff, segment->size);
+ } else {
+ memcpy(data + offset, ptr, segment->size);
+ }
--
Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc, is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
