On Tue, 2023-09-12 at 15:20 -0400, Antoni Boucher wrote:
FWIW I've pushed the "ggc, jit: forcibly clear GTY roots in jit" to
trunk after retesting it, as r14-4003-geaa8e8541349df.
> I added it to bugzilla here:
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111396
I don't yet have a fix for this
I added it to bugzilla here:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111396
Since this only reproduces part of the issue, please let me test again
with rustc_codegen_gcc after adding the missing fix.
I confirmed that the fix is in
https://github.com/antoyo/gcc/commit/9d5b6b20efa2082592619675
On Tue, 2023-09-12 at 13:36 -0400, Antoni Boucher wrote:
> In the mean time, here's a (Rust) reproducer for the issue:
>
> fn main() {
> for _ in 0..5 {
> let context = Context::default();
> context.add_command_line_option("-flto");
>
> context.set_optimization_level(Op
In the mean time, here's a (Rust) reproducer for the issue:
fn main() {
for _ in 0..5 {
let context = Context::default();
context.add_command_line_option("-flto");
context.set_optimization_level(OptimizationLevel::Aggressive);
context.add_driver_option("-nostdli
It seems to not be enough to fix the issue.
Let me find out what's missing from my patch.
On Tue, 2023-09-12 at 11:35 +0200, Richard Biener via Jit wrote:
> On Wed, Sep 6, 2023 at 3:41 PM David Malcolm via Gcc-patches
> wrote:
> >
> > As part of Antoyo's work on supporting LTO in rustc_codegen_g
On Wed, Sep 6, 2023 at 3:41 PM David Malcolm via Gcc-patches
wrote:
>
> As part of Antoyo's work on supporting LTO in rustc_codegen_gcc, he
> noticed an ICE inside libgccjit when compiling certain rust files.
>
> Debugging libgccjit showed that outdated information from a previous
> in-memory comp
Hi.
I'll do another test to make sure this is enough since I tested with a
few more finalize functions.
Thanks a lot for finding this!
On Wed, 2023-09-06 at 09:40 -0400, David Malcolm via Jit wrote:
> As part of Antoyo's work on supporting LTO in rustc_codegen_gcc, he
> noticed an ICE inside libgc
As part of Antoyo's work on supporting LTO in rustc_codegen_gcc, he
noticed an ICE inside libgccjit when compiling certain rust files.
Debugging libgccjit showed that outdated information from a previous
in-memory compile was referring to ad-hoc locations in the previous
compile's line_table.
The
