Re: [PATCH] net: wireless: iwlegacy: Fix possible data races in il4965_send_rxon_assoc()

On 2018/10/5 15:54, Stanislaw Gruszka wrote:

On Thu, Oct 04, 2018 at 04:52:19PM +0800, Jia-Ju Bai wrote:

quoted

On 2018/10/4 15:59, Stanislaw Gruszka wrote:

quoted

On Wed, Oct 03, 2018 at 10:07:45PM +0800, Jia-Ju Bai wrote:

quoted

These possible races are detected by a runtime testing.
To fix these races, the mutex lock is used in il4965_send_rxon_assoc()
to protect the data.

Really ? I'm surprised by that, see below.

My runtime testing shows that il4965_send_rxon_assoc() and
il4965_configure_filter() are concurrently executed.
But after seeing your reply, I need to carefully check whether my
runtime testing is right, because I think you are right.
In fact, I only monitored the iwl4965 driver, but did not monitor
the iwlegacy driver, so I will do the testing again with monitoring
the lwlegacy driver.

<snip>

quoted

quoted

So I wonder how this patch did not cause the deadlock ?

Oh, sorry, anyway, my patch will cause double locks...

So how those runtime test were performend such you didn't
notice this ?

I write a tool to perform runtime testing.
This tool records the lock status during driver execution.
Some calls to mutex_lock() are in common.c that I did not handle, so the 
corresponding lock status was not recorded by my tool, causing this 
false positive.

Now I have handled common.c, and this false positive is not reported any 
more.
Actually, I get several new reports.
I will send you these reports to you later, and hope you can have a 
look, thanks in advance :)

quoted

quoted

Anyway what can be done is adding:

lockdep_assert_held(&il->mutex);

il4965_commit_rxon() to check if we hold the mutex.

I agree.

Care to post a patch ?

Sure :)


Best wishes,
Jia-Ju Bai