On 04/10/17 00:50, [email protected] wrote:
> From: Vladimir Mezentsev <[email protected]>
>
> Tested on aarch64-linux-gnu.
> No regression.
> No bootstrap failure.
>
> gcc/ChangeLog:
> 2017-09-26 Vladimir Mezentsev <[email protected]>
>
> * gcc/config/aarch64/aarch64.c: restore fix in
> aarch64_use_blocks_for_constant_p
Vladimir,
Thanks for the patch, but I can't accept this as it stands. The patch
lacks a description of what is motivating the change, so I can't really
review it sensibly. What's the problem you are trying to address? Why
do you believe this is the right fix? Do you have a test case?
The ChangeLog entry doesn't help me either: restore what fix?
Is the patch associated with a bugzilla report? If so, then please
include that in the ChangeLog in the form "PR <category>/<pr-num>.
Incidentally, ChangeLog entries should be complete sentences (begin with
a capital letter and end with a full stop).
R.
> ---
> gcc/config/aarch64/aarch64.c | 8 +++-----
> 1 file changed, 3 insertions(+), 5 deletions(-)
>
> diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
> index 1c14008..b377bc7 100644
> --- a/gcc/config/aarch64/aarch64.c
> +++ b/gcc/config/aarch64/aarch64.c
> @@ -6193,11 +6193,9 @@ aarch64_can_use_per_function_literal_pools_p (void)
> static bool
> aarch64_use_blocks_for_constant_p (machine_mode, const_rtx)
> {
> - /* Fixme:: In an ideal world this would work similar
> - to the logic in aarch64_select_rtx_section but this
> - breaks bootstrap in gcc go. For now we workaround
> - this by returning false here. */
> - return false;
> + /* We can't use blocks for constants when we're using a per-function
> + constant pool. */
> + return !aarch64_can_use_per_function_literal_pools_p ();
> }
>
> /* Select appropriate section for constants depending
>
