This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc 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: Remove -fexceptions configure test


On 19/10/15 17:44, Joseph Myers wrote:
On Sun, 18 Oct 2015, Szabolcs Nagy wrote:

i think -fasynchronous-unwind-tables don't need -fexceptions
(currently).

-fasync* means unwind can propagate from anywhere, cleanups
are not run.

So do you think cases with callbacks (which might be C++ functions
throwing exceptions through libc code) need -fexceptions, or only
-funwind-tables?


i think -funwind-tables is enough if there is no
__attribute__((cleanup(foo))) usage in the code
around the callback.

it makes sense to use -fexceptions consistently
anyway to be safe and future proof.

i just wanted to note the async case because it
is going to change with the cancel redesign so the
patch should take that into consideration.
(probably async unwind will not be needed at all then,
so -fexceptions can be used consistently for callbacks
and cancellation points).


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