Re: [PR51752] publication safety violations in loop invariant motion pass

[Andrew MacLeod <amacleod@redhat.com>]

On 02/28/2012 04:39 AM, Richard Guenther wrote:
> On Mon, Feb 27, 2012 at 5:44 PM, Andrew MacLeod<amacleod@redhat.com>  wrote:
> I t would seem appropriate to me that in the future, perhaps the CFG could
> have a flag set for any basic block which is in a transaction...  This would
> make it pretty trivial at all times to tell if a statement is part of a
> transaction or not.  It seems like the process of CFG construction/updating
> should be able to maintain that info for little cost. (ie, I'd guess it's
> already making the traversals requires to collect that info)
>
> It would certainly be convenient :-)
> That would sound easier than maintaining the flag on the stmt.  OTOH we
> already maintain on-the-side EH tables, so it can't be that big of a problem.
>
> What do we do if we move stmts inside a transaction though?  Do they
> magically become part of it and thus we cannot move them out again?
well, I think by definition if you move a statement into a basic block 
of a transaction it becomes a part of the transaction.  I believe 
transactions are block based... the entire block is in f a transaction, 
or not.    Moving a statement out of a transaction would be as easy as 
moving it in... just change the block it is in.

> If you want to fix this issue for 4.7, add the gimple flag and an on-demand
> compute helper and check the flag in PRE/LIM.
>
>
yes, a targeted approach is required for 4.7.  I just wanted to plant 
the seed while the discussion was ongoing.

Andrew