https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Andrew Pinski changed:
What|Removed |Added
CC||user202729 at protonmail dot
com
---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Jorg Brown changed:
What|Removed |Added
CC||jorg.brown at gmail dot com
--- Comment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #14 from Richard Biener ---
I think implementation-wise GCC outrules aliases that are not visible but takes
care of symbols resolving to NULL. For optimizations of actual accesses it can
assume the symbols do not resolve to NULL
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #13 from Martin Sebor ---
As noted in the duplicate pr90122, the test case below shows that GCC already
relies on different extern declarations denoting distinct objects. It just
doesn't fold the address equality expression for some
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #12 from Martin Sebor ---
*** Bug 90122 has been marked as a duplicate of this bug. ***
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #11 from Marc Nieper-Wißkirchen ---
If ISO C allows such linkage to be created outside of the standard, a number of
other assumption would be violated as well:
In 6.2.4 (2) it says that "an object exists, has a constant address, and
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #10 from joseph at codesourcery dot com ---
My interpretation of that footnote is that it's observing that there is no
way within the standard to *create* linkage between different identifiers
- not that it constrains how such
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #9 from Marc Nieper-Wißkirchen ---
Footnote 29) in section 6.2.2 of the latest draft (N2176) for C18 says: "There
is no linkage between different identifiers."
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #8 from rguenther at suse dot de ---
On February 1, 2019 4:38:29 PM GMT+01:00, "jakub at gcc dot gnu.org"
wrote:
>https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
>
>Jakub Jelinek changed:
>
> What|Removed
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #7 from Marc Nieper-Wißkirchen ---
I'm sorry, I wasn't precise what I meant. When I wrote that the optimization
wouldn't be possible I meant the case of two externally defined variables, e.g.
extern int p;
extern int q;
One can
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #6 from Jakub Jelinek ---
That optimization is of course possible if the compiler can prove the addresses
are different. So, e.g. if one of the two vars is defined locally, or both
locally, or one is automatic and another
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
--- Comment #5 from Marc Nieper-Wißkirchen ---
If that was possible (that symbols are aliased in the TU in which they are
defined, but not (explicitly) in a TU where they are declared), there would be
the need of a "no_alias" (or "never_alias")
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Jakub Jelinek changed:
What|Removed |Added
CC||jakub at gcc dot gnu.org
--- Comment #4
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Richard Biener changed:
What|Removed |Added
CC||m...@nieper-wisskirchen.de
--- Comment
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Richard Biener changed:
What|Removed |Added
CC||hubicka at gcc dot gnu.org,
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=43565
Martin Sebor changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Last reconfirmed|
16 matches
Mail list logo
