On 9/12/22 6:25 PM, Jeff Johnson wrote:
Change ff6d365898d ("soc: qcom: qmi: use const for struct qmi_elem_info") allows QMI message encoding/decoding rules to be const. So now update the definitions in the various client to take advantage of this. Patches for ath10k and ath11k were perviously sent separately.
I have had this on my "to-do list" for ages. The commit you mention updates the code to be explicit about not modifying this data, which is great.
I scanned over the changes, and I assume that all you did was make every object having the qmi_elem_info structure type be defined as constant.
Why aren't you changing the "ei_array" field in the qmi_elem_info structure to be const? Or the "ei" field of the qmi_msg_handler structure? And the qmi_response_type_v01_ei array (and so on)?
I like what you're doing, but can you comment on what your plans are beyond this series? Do you intend to make the rest of these fields const?
Thanks.
-Alex
This series depends upon: https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux.git/commit/?h=for...
This is in the for-next banch of: git.kernel.org/pub/scm/linux/kernel/git/qcom/linux.git
Hence this series is also based upon that tree/branch.
Jeff Johnson (4): net: ipa: Make QMI message rules const remoteproc: sysmon: Make QMI message rules const slimbus: qcom-ngd-ctrl: Make QMI message rules const soc: qcom: pdr: Make QMI message rules const
drivers/net/ipa/ipa_qmi_msg.c | 20 ++++++++++---------- drivers/net/ipa/ipa_qmi_msg.h | 20 ++++++++++---------- drivers/remoteproc/qcom_sysmon.c | 8 ++++---- drivers/slimbus/qcom-ngd-ctrl.c | 8 ++++---- drivers/soc/qcom/pdr_internal.h | 20 ++++++++++---------- 5 files changed, 38 insertions(+), 38 deletions(-)