Hi Peter, > > The best solution would be to pass the buffer size as an extra > parameter. Especially when some code passes buffers that are > allocated/reserved dynamically. > > Sigh, I am not sure how many changes it would require in kallsyms > API and all the callers. But it would be really appreciated, IMHO. >
yes we already prepared size changes 5-6 months back: https://lore.kernel.org/lkml/yontol4zc4cyt...@infradead.org/t/ [PATCH 1/5] kallsyms: pass buffer size in sprint_* APIs But at that time new API development(for replacement of seq_buf) was in progress and we decided to wait for that completion. https://lore.kernel.org/r/20220604193042.1674951-2-kent.overstr...@gmail.com https://lore.kernel.org/r/20220604193042.1674951-4-kent.overstr...@gmail.com As I checeked these APIs are not pushed to mainline. we will try to prepare new patch set for kallsym changes again with seq_buf to take care of length argument. Thanks, Maninder Singh
