[PATCH 1/5] clk: allow reentrant calls into the clk framework

On Thu, 2013-02-28 at 12:49 +0800, Mike Turquette wrote:

Reentrancy into the clock framework from the clk.h api is highly
desirable.  This feature is necessary for clocks that are prepared and
unprepared via i2c_transfer (which includes many PMICs and discrete
audio chips) and it is also necessary for performing dynamic voltage &
frequency scaling via clock rate-change notifiers.

This patch implements reentrancy by adding a global atomic_t which
tracks the context of the current caller.  Context in this case is the
return value from get_current().  The clk.h api implementations are
modified to first see if the relevant global lock is already held and if
so compare the global context (set by whoever is holding the lock)
against their own context (via a call to get_current()).  If the two
match then this function is a nested call from the one already holding
the lock and we procede.  If the context does not match then procede to
call mutex_lock and busy-wait for the existing task to complete.

Thus this patch set does not increase concurrency for unrelated calls
into the clock framework.  Instead it simply allows reentrancy by the
single task which is currently holding the global clock framework lock.

Thanks to Rajagoapl Venkat for the original idea to use get_current()
and to David Brown for the suggestion to replace my previous rwlock
scheme with atomic operations during code review at ELC 2013.

Signed-off-by: Mike Turquette <redacted>
Cc: Rajagopal Venkat <redacted>
Cc: David Brown <redacted>
---

Hi Mike,

Will this single patch be accepted? I guess you might not merge the
whole series but I think this one is useful, is it possible that you can
send out this single patch (or just merge this one) as an improvement of
CCF? Or you think otherwise?

Thanks,
Bill