Re: [PATCH v4 7/9] mm/vmscan: pass scan_control to mm_vmscan_balance_pgdat_end
From: Ridong Chen
Date: Sun Oct 25 2026 - 03:26:35 EST
On 10/25/2026 6:24 AM, Barry Song wrote:
On Mon, Oct 19, 2026 at 12:24 PM Ridong Chen <ridong.chen@xxxxxxxxx> wrote:Sure, I split the patches just to make them easier to review, since each one contains only small changes.
From: Ridong Chen <chenridong@xxxxxxxxxx>
balance_pgdat() passes sc.order and sc.nr_reclaimed to
mm_vmscan_balance_pgdat_end as bare arguments. Pass the enclosing
struct scan_control instead and read the fields out in TP_fast_assign,
consistent with the other reclaim tracepoints. highest_zoneidx stays a
separate argument: it is a balance_pgdat() parameter, and
sc.reclaim_idx may differ from it by the time the end event fires.
No functional change. The tracefs event format (TP_STRUCT__entry and
TP_printk) is unchanged, so tracefs-based consumers are unaffected.
BPF programs attached via raw_tp/tp_btf see the new TP_PROTO and need
to be adjusted; tracepoints are not a stable ABI (see
Documentation/bpf/bpf_design_QA.rst).
Assisted-by: LLM
Signed-off-by: Ridong Chen <chenridong@xxxxxxxxxx>
Could we combine patches 4–7 into a single patch? The changes
seem a bit too scattered across separate patches.
--
Best regards
Ridong