[RFC PATCH 0/8] zram: introduce multi-handle entries

From: Sergey Senozhatsky
Date: Mon Nov 18 2024 - 05:31:14 EST


*** RFC ***

: Given that it's merge window time, consider this low priority.

ZRAM_HUGE objects are incompressible and each takes a whole
physical page on the zsmalloc side. zsmalloc pool, natually, has
some internal memory fragmentation (within size-classes) and what
we can do for ZRAM_HUGE objects is to split them into several
smaller objects (2 at this point) and store those parts individually
in regular size-classes (hence multi-handle entries). This, basically,
lets us to use already allocated (but unused) zspage memory for
ZRAM_HUGE objects, instead of unconditional allocation of 0-order
page for each ZRAM_HUGE object.

Sergey Senozhatsky (8):
zram: cond_resched() in writeback loop
zram: free slot memory early during write
zram: remove entry element member
zram: handle ZRAM_SAME immediately
zram: factor out ZRAM_HUGE write
zram: factor out ZRAM_HUGE read
zsmalloc: move ZS_HANDLE_SIZE to zmslloc header
zram: introduce multi-handle entries

drivers/block/zram/zram_drv.c | 351 +++++++++++++++++++++-------------
drivers/block/zram/zram_drv.h | 12 +-
include/linux/zsmalloc.h | 2 +
mm/zsmalloc.c | 2 -
4 files changed, 236 insertions(+), 131 deletions(-)

--
2.47.0.338.g60cca15819-goog