Re: [PATCH] PR91195: fix -Wmaybe-uninitialized warning for conditional store optimization

[Jeff Law <law@redhat.com>]

On 7/24/19 12:07 PM, Martin Sebor wrote:

> 
> I don't know what Jakub had in mind but the mapping I envision is
> one like hash_map<tree, bitmap> that would make it possible to set
> a bit for each distinct warning for every tree node.  It would let
> us set a bit for -Wuninitialized while leaving the bit for
> -Warray-bounds (and all other warnings) clear.
Ah, yes.  I like that.  I'm still worried about the linkage between the
map and the GC system, but a <tree, bitmap> has a lot of potential.

> 
>>
>> If the bit were on an SSA_NAME, or a _DECL node, then the flag bit is
>> shared and would be a much larger concern.
> 
> For shared objects the mapping would have to be more involved but
> I haven't thought about it in any detail to have an idea what it
> might look like.
I suspect shared objects are just going to be painful.  A solution which
worked on EXPR nodes would still be a significant step forward.


> 
> Anyway, if/when someone does come up with a solution for this we
> will have to go through all the places where the no-warning bit
> is set and replace them with whatever replacement we come up with.
> One instance more or less won't make a difference.  I just wanted
> to point out that setting the bit is not a robust solution.
Yea, but at least they're easy to find via the TREE_NO_WARNING flag.
Hopefully we've got tests for most of the issues we're working around
with that bit.

jeff