PERF-IOSTAT(1)
π
PERF-IOSTAT(1) β the perf iostat glow-up π
okay bestie, so this is perf iostat, and it lives in the perf manual (yes, that perf β¨), and honestly? it's giving "I want to know exactly how much data my PCIe root ports are slinging around" and ngl it ate. π₯
β¨ NAME β¨
perf-iostat β show I/O performance metrics π¦
πΈ SYNOPSIS πΈ
perf iostat list
perf iostat <ports> -- <command> [<options>]
π DESCRIPTION π
so this mode is literally designed to hand you four I/O performance metrics per each PCIe root port. four. metrics. per port. periodt. π
- Inbound Read β I/O devices below the root port read from the host memory, in MB
- Inbound Write β I/O devices below the root port write to the host memory, in MB
- Outbound Read β CPU reads from I/O devices below the root port, in MB
- Outbound Write β CPU writes to I/O devices below the root port, in MB
like, that's the whole vibe check: who's reading, who's writing, and which direction the bytes are going. π«Ά
π OPTIONS π
<command>...β literally any command you can specify in a shell, bestie. go wild. π§listβ list all the PCIe root ports. all of them. π<ports>β select which root ports you want to monitor. a comma-separated list is supported, so you can pick a lil squad instead of the whole group chat. π
β EXAMPLES β
1. list all PCIe root ports (example for 2-S platform) π
$ perf iostat list
S0-uncore_iio_0<0000:00>
S1-uncore_iio_0<0000:80>
S0-uncore_iio_1<0000:17>
S1-uncore_iio_1<0000:85>
S0-uncore_iio_2<0000:3a>
S1-uncore_iio_2<0000:ae>
S0-uncore_iio_3<0000:5d>
S1-uncore_iio_3<0000:d7>
look at them all lined up. iconic behavior honestly. β¨
2. collect metrics for all PCIe root ports π₯
$ perf iostat -- dd if=/dev/zero of=/dev/nvme0n1 bs=1M oflag=direct
357708+0 records in
357707+0 records out
375083606016 bytes (375 GB, 349 GiB) copied, 215.974 s, 1.7 GB/s
Performance counter stats for 'system wide':
port Inbound Read(MB) Inbound Write(MB) Outbound Read(MB) Outbound Write(MB)
0000:00 1 0 2 3
0000:80 0 0 0 0
0000:17 352552 43 0 21
0000:85 0 0 0 0
0000:3a 3 0 0 0
0000:ae 0 0 0 0
0000:5d 0 0 0 0
0000:d7 0 0 0 0
not the 0000:17 port absolutely carrying the entire team with 352552 inbound read MB π main character energy fr fr.
3. collect metrics for a comma-separated list of PCIe root ports π«
$ perf iostat 0000:17,0:3a -- dd if=/dev/zero of=/dev/nvme0n1 bs=1M oflag=direct
357708+0 records in
357707+0 records out
375083606016 bytes (375 GB, 349 GiB) copied, 197.08 s, 1.9 GB/s
Performance counter stats for 'system wide':
port Inbound Read(MB) Inbound Write(MB) Outbound Read(MB) Outbound Write(MB)
0000:17 358559 44 0 22
0000:3a 3 2 0 0
197.081983474 seconds time elapsed
see how we only invited two ports to the party? that's the comma-separated list doing its thing. so slay. π
π¦ SEE ALSO π¦
perf-stat(1)β go check out the sibling, iykyk π
lowkey reminder: keep<ports>comma-separated and don't forget the--before your<command>and[<options>], bestie. the vibes depend on it. π
β¨ that's the whole page! β¨ this gem is from the perf manual, dated 10/31/2026 β written by perf, and it's PERF-IOSTAT(1) all day every day. go collect some metrics, you absolute legend πππΈ
π¦ where this page comes from
- arch rolling: perf 7.3.1-1 Β· GPL-2.0-only
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms