This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Fix PR 24670. Prune equivalence sets in ASSERT_EXPR chains


> 
> 
> Description of the problem is in the code.  Essentially, given a chain of 
> ASSERT_EXPRs we cannot always add the old names in the chain to the 
> equivalence set of the new one.  Sometimes the assertions are 
> incompatible.


A real good question is the Why are we so worried about dead code
and abort because the asserts are incompatible?

>From the looks of things, you increased compile time for cases like
this.

Now we have an O(n) loop which does nothing really except make sure we
are not aborting.  I don't get it.

-- Pinski


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]