linux: add a disk-backed swap meter

[blacks1ne]

Expose disk-backed swap separately from zram so users can distinguish backing-store pressure from compressed RAM usage.
I chose Dsw abbreviation that is open to a better alternative.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: d1a117eb-402d-4316-aec0-d464be0d8292

📥 Commits

Reviewing files that changed from the base of the PR and between 82b0400 and ef24653.

📒 Files selected for processing (1)

• linux/DiskSwapMeter.c

Included review availability: Your plan provides up to 8 included reviews per hour; 5 remain after this review.

---

📝 Walkthrough

Walkthrough

The change collects non-zram disk-backed swap usage from /proc/swaps. It stores total and used values in LinuxMachine. A new DiskSwapMeter formats and displays these values. The platform registers the meter and transfers machine values to it. The new header and source files are added to the Linux build lists.

Poem

Swap pages gather, zram steps aside,
Procfs records the values supplied.
A meter shows totals in view,
The platform carries the numbers through.
Build lists bind the feature tight.

Merge Risk: ⚪ Minimal · up to ef246

No actionable merge-blocking risk remains; the change is merge-ready after normal checks and review.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[fasterit]

There's a swap meter already. Extend that in case you feel the zram use case warrants that?

[blacks1ne]

There's a swap meter already. Extend that in case you feel the zram use case warrants that?

Thanks, I'm ready to extend the existing Swap meter rather than add a separate meter.

One UI question before I revise it.

/proc/swaps provides per-device Used values, but those include pages reported globally as SwapCached. The current Swap meter shows that cache separately.

Would you prefer disk-backed and zram components to show backing-store occupancy, with SwapCached moved to text-only/removed from the bar.

That would make the displayed disk/zram components honest and non-overlapping.

Let me try to illustrate:

Submitted implementation

Swap: 2.25G / 40G
[ used 2.25G ][ cache 0.25G ][ free … ]

Zram: 0.7G (2.0G) / 8G
[ compressed 0.7G ][ uncompressed 1.3G ][ free … ]

Suggested: extend the Swap meter

Swap: 2.5G / 40G cache: 0.25G (included in a backing store)
[ Disk 0.5G ][ Zram 2.0G ][ free … ]

[fasterit]

Do the math properly?

ChatGPT Codex recommends:

/proc/swaps gives you enough to calculate current swap usage by device, including zram.

Example:

$ cat /proc/swaps

Filename                Type        Size        Used        Priority

/dev/zram0              partition   8388604     1234567     100

/dev/sda3               partition   16777212    2345678     -2

The columns are in KiB:

• Size = total swap capacity

• Used = currently used swap

• Free = Size - Used

• Usage % = Used / Size × 100

So in this example:

Device	Type	Size	Used	Usage
/dev/zram0	zram	8 GiB	~1.18 GiB	~14.7%
/dev/sda3	disk	16 GiB	~2.24 GiB	~13.9%

Shell command

You can calculate it directly from /proc/swaps:

awk 'NR>1 {
    size=$3
    used=$4
    printf "%-20s size=%6.2f GiB used=%6.2f GiB free=%6.2f GiB usage=%5.1f%%\n",
           $1, size/1048576, used/1048576,
           (size-used)/1048576, 100*used/size
}' /proc/swaps

To specifically distinguish zram vs disk swap, you can use:

awk 'NR>1 {
    if ($1 ~ /zram/) zram_size+=$3; zram_used+=$4
    else disk_size+=$3; disk_used+=$4
}
END {
    printf "zram: %6.2f GiB used / %6.2f GiB total (%5.1f%%)\n",
           zram_used/1048576, zram_size/1048576,
           100*zram_used/zram_size
    printf "disk: %6.2f GiB used / %6.2f GiB total (%5.1f%%)\n",
           disk_used/1048576, disk_size/1048576,
           disk_size ? 100*disk_used/disk_size : 0
}' /proc/swaps

Important: zram's Used in /proc/swaps is the amount of compressed data occupying zram swap, not necessarily the amount of physical RAM consumed by the zram device. If you're trying to calculate the actual RAM overhead of zram, /sys/block/zram*/mm_stat is more informative.

[blacks1ne]

Thank you for the review and guidance.

I’ll maintain the change in my fork as a separate disk-swap meter. Its purpose is to make disk-backed swap immediately visible, since it has materially different latency and stall implications from RAM-backed zram, a direction that may not fit upstream.

The existing Zram meter remains useful for its compression and RAM-cost view.