> ## Documentation Index
> Fetch the complete documentation index at: https://docs.jylhis.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Performance

> Performance tuning options

## CPU mitigations

| Option                                   | Type   | Default | Description                                                   |
| ---------------------------------------- | ------ | ------- | ------------------------------------------------------------- |
| `marchyo.performance.disableMitigations` | `bool` | `true`  | Disable CPU vulnerability mitigations for maximum performance |

```nix theme={}
marchyo.performance.disableMitigations = true;
```

<Warning>
  Disabling CPU mitigations (Spectre, Meltdown, etc.) reduces security. Only enable on trusted single-user workstations where maximum performance is required (e.g., gaming, benchmarking). Do **not** enable if running untrusted code or containers.
</Warning>

To keep mitigations enabled:

```nix theme={}
marchyo.performance.disableMitigations = false;
```

## Kernel & I/O tuning

Opt-in kernel/sysctl/IO tuning for throughput-oriented workloads. Disabled by
default. Enabling `marchyo.performance.tuning.enable` turns on the broadly-safe
sub-toggles (`network`, `nvme`, `memory`); the aggressive toggles (`hugePages`,
`compute`) stay off and must be opted into explicitly.

| Option                                                   | Type   | Default          | Description                                                                 |
| -------------------------------------------------------- | ------ | ---------------- | --------------------------------------------------------------------------- |
| `marchyo.performance.tuning.enable`                      | `bool` | `false`          | Master switch; enables the safe sub-toggles below                           |
| `marchyo.performance.tuning.network.enable`              | `bool` | follows `enable` | BBR congestion control, TCP Fast Open, MTU probing, larger socket buffers   |
| `marchyo.performance.tuning.nvme.enable`                 | `bool` | follows `enable` | NVMe udev tuning: no-op scheduler, larger read-ahead and max request size   |
| `marchyo.performance.tuning.memory.enable`               | `bool` | follows `enable` | Lower swappiness, keep FS metadata cached, byte-based dirty-page thresholds |
| `marchyo.performance.tuning.memory.dirtyBytes`           | `int`  | `268435456`      | `vm.dirty_bytes` (256 MiB) — synchronous writeback threshold                |
| `marchyo.performance.tuning.memory.dirtyBackgroundBytes` | `int`  | `67108864`       | `vm.dirty_background_bytes` (64 MiB) — background writeback threshold       |
| `marchyo.performance.tuning.hugePages.enable`            | `bool` | `false`          | 2 MiB transparent huge pages always                                         |
| `marchyo.performance.tuning.compute.enable`              | `bool` | `false`          | Relaxed PAM limits (memlock/rtprio/nofile/nproc) for trusted compute hosts  |

```nix theme={}
# Safe defaults (network + nvme + memory):
marchyo.performance.tuning.enable = true;

# Compute/CUDA workstation — opt into the aggressive toggles too:
marchyo.performance.tuning = {
  enable = true;
  hugePages.enable = true;
  compute.enable = true;
};
```

<Warning>
  `hugePages` can hurt latency-sensitive and interactive desktop workloads.
  `compute` relaxes resource limits system-wide (unlimited `memlock`, real-time
  priority for all users) — only enable it on trusted single-user machines where
  you control every process.
</Warning>

<Note>
  The CFS scheduler sysctls from older compute-tuning sets
  (`kernel.sched_min_granularity_ns`, `sched_latency_ns`, etc.) are intentionally
  **not** included: they were removed when the kernel switched CFS → EEVDF (6.6+)
  and setting them on a current kernel only produces `systemd-sysctl` warnings.
</Note>