More comment improvements for RCU primitives
A message that is not in the archive (sent off-list or not yet received): [email protected]
A message that is not in the archive (sent off-list or not yet received): 20040722112520.6b7a21b9@dell_ss3.pdx.osdl.net
Quoted 5 lines · Stephen Hemminger
Quoted 2 lines · On Mon, 19 Jul 2004 19:35:15 -0700 Paul Mckenney <[email protected]>…
Hello!
Quoted 8 lines
Quoted 5 lines
Here is a patch to improve the usefulness of the RCU primitives' documentation. Again, this probably interacts badly with existing RCU patches, which I will fix when I incorporate feedback.
Thanx, Paul
If you are going to be this verbose (which is good), you may want to mention how this interacts on a UP system as well.
Good point...
My first thought would be to add a paragraph saying what happens on a UP for call_rcu() and synchronize_kernel(), something like for call_rcu():
This primitive has the same effect on a UP system. For example, if an RCU read-side critical section is interrupted by a handler that invokes call_rcu(), the corresponding update function will be invoked some time after the RCU read-side critical section has completed, which will in turn be some time after the interrupt handler returns.
Does this seem reasonable?
Thanx, Paul