Misc #14764: [PATCH] vm_trace: remove rb_postponed_job_t->flags

• Status: Closed
• Assignee: -

[normalperson (Eric Wong)]

vm_trace: remove rb_postponed_job_t->flags

This field has been taking up space for 5 years and never used.
Save 8kb of memory on x86-64 for now.  If we ever care about
this flag, we'll simply re-add it.

No ABI or API compatibility changes.

* vm_trace.c (rb_postponed_job_struct): remove flags
  (postponed_job_register): comment out flags assignment

Seems trivial, will commit soon unless somebody decides to
start using this flag now (after 5 years).

[normalperson (Eric Wong)]

Btw, 1000 slots for postponed jobs seems like overkill,
as does the use of atomics.

I think we can use a st_table with st_shift...
Hash key can be the func pointer, since this API deduplicates
calls based on function.

[ko1 (Koichi Sasada)]

no problem. st_table also seems good.

[normalperson (Eric Wong)]

[e-mail] wrote:

no problem. st_table also seems good.

OK, this uses st_table:
https://80x24.org/spew/[e-mail]/raw
(only lightly-tested)

[normalperson (Eric Wong)]

• Status changed from Open to Closed

Applied in changeset trunk|r63450.

---

vm_trace: remove rb_postponed_job_t->flags

This field has been taking up space for 5 years and never used.
Save 8kb of memory on x86-64 for now. If we ever care about
this flag, we'll simply re-add it.

• vm_trace.c (rb_postponed_job_struct): remove flags
  (postponed_job_register): comment out flags assignment
  [ruby-core:87052] [Misc #14764]