On 6/19/25 14:48, chuguangqing wrote:
tks to pabeni. cc the relevant ML in the email.
Thank you for reaching out with the patch, it is an attempt to fix something observable? I have found the original topic via web search [1]. It looks that the proposed patch does not change anything - adapter->io_addr was null also before the assignment, so it does not hurt as-is. And the check afterwards is equally effective for me to prevent subsequent dereference. [1] https://lkml.org/lkml/2025/6/19/343
