Re: make dobench failing recently for me

[Charles Greathouse <crgreathouse@gmail.com>]

That cleared up all but the number-dyn.dif issue, which is still having heap issues:

cat Olinux-x86_64/number-dyn.dif
*** ../src/test/32/number       2023-01-26 20:24:02.486620100 -0500
--- gp.out      2024-01-22 14:45:42.896234900 -0500
***************
*** 276,280 ****
  [768, [12, 4, 4, 2, 2], [Mod(2641, 3120), Mod(2497, 3120), Mod(2341, 3120),
  Mod(1951, 3120), Mod(2081, 3120)]]
  ? if(getheap()!=HEAP,getheap())
  ? print("Total time spent: ",gettime);
! Total time spent: 13
--- 276,281 ----
  [768, [12, 4, 4, 2, 2], [Mod(2641, 3120), Mod(2497, 3120), Mod(2341, 3120),
  Mod(1951, 3120), Mod(2081, 3120)]]
  ? if(getheap()!=HEAP,getheap())
+ [96, 3256]
  ? print("Total time spent: ",gettime);
! Total time spent: 16

On Mon, Jan 22, 2024 at 1:14 PM Bill Allombert <Bill.Allombert@math.u-bordeaux.fr> wrote:

On Mon, Jan 22, 2024 at 01:04:12PM -0500, Charles Greathouse wrote:
> I'm not sure if this is a bug or local configuration issues, so I'll post
> here for now. While I'm able to build gp from the repo, for a while (weeks,
> at least) I haven't been able to pass the bench. The diffs show a
> precision/rounding issue in trans-dyn and heap/memory leak issues in
> analyz-dyn, number-dyn, and polyser-dyn.

Most probably this is caused by ./Configure --tune
Do you have a file Olinux-x86_64/tune.h ?

In that case you probably need to update it.

Cheers,
Bill.