Hi Sourabh, Christophe, Thanks for checking this.
I rechecked the current sysfs write path and I no longer think this patch is valid. sysfs_kf_write() returns 0 before calling ->store() when count is 0, so I do not have a path that reaches bootargs_append_store() with count == 0. Also, if such a path did exist, returning 0 would be the right behavior for an empty write, not -EINVAL. I'll drop this patch. Thanks, Pengpeng
