softirqs(8)

🦋 NAME 🦋

okay so softirqs is the tool that measures soft IRQ (soft interrupt) event time 🧚 It uses Linux eBPF/bcc, so it's giving low-level, and I'm obsessed 💅

💖 SYNOPSIS 💖

softirqs [-h] [-T] [-N] [-C] [-d] [-c CPU] [interval] [count]

✨ DESCRIPTION ✨

bestie, this one summarizes the time spent servicing soft IRQs (soft interrupts) 🫶 You can see that time as plain totals or as histogram distributions, your choice 📊 A system-wide version of this time lives in the %soft column of mpstat(1), and soft IRQ event counts (but not times) are chilling in /proc/softirqs 🌸

This tool uses the irq:softirq_enter and irq:softirq_exit kernel tracepoints, which is a stable tracing mechanism, so no notes 💜 BPF programs can attach to tracepoints from Linux 4.7 only, so keep that in mind 🙈 An older version is in tools/old, and it's kprobes-based instead of tracepoint-based. Very retro 📼

Since this uses BPF, only the root user can use this tool 🔒 Sorry not sorry, that's the rule.

🎀 REQUIREMENTS 🎀

You need CONFIG_BPF and bcc. No way around it 💀

👑 OPTIONS 👑

• -h — pulls up the usage message 📖
• -T — slaps timestamps onto the output ⏰
• -N — outputs in nanoseconds 🔬
• -C — shows the number of soft irq events 🔢
• -d — shows the IRQ time distribution as histograms 📊
• -c CPU — traces on this CPU only, the one you pick 🎯

🌈 EXAMPLES 🌈

Sum the soft IRQ event time until you hit Ctrl-C:

# softirqs

Show the number of soft irq events:

# softirqs -C

Show soft IRQ event time as histograms:

# softirqs -d

Print 1 second summaries, 10 times:

# softirqs 1 10

1 second summaries, printed in nanoseconds, with timestamps:

# softirqs -NT 1

Sum soft IRQ event time on CPU 1 until Ctrl-C:

# softirqs -c 1

🍓 FIELDS 🍓

• SOFTIRQ — the kernel function name doing the soft IRQ action 🛠️
• TOTAL_usecs — total time spent in this soft IRQ function, in microseconds ⏱️
• TOTAL_nsecs — total time spent in this soft IRQ function, in nanoseconds ⏱️
• usecs — the microsecond range for this bucket 📏
• nsecs — the nanosecond range for this bucket 📏
• count — how many soft IRQs landed in this time range 🧮
• distribution — an ASCII picture of the distribution (the count column) 📈

🔥 OVERHEAD 🔥

okay real talk, this traces kernel functions and keeps in-kernel counts, which get asynchronously copied to user-space. If the interrupt rate is very high (>1M/sec), it's still a relatively efficient way to trace these events, so overhead should be small for normal workloads 🥹 It could get noticeable for heavy workloads though. Measure in a test environment before use 🧪

🌸 SOURCE 🌸

This is from bcc, obviously 💜

https://github.com/iovisor/bcc

Also look in the bcc distribution for a companion _examples.txt file with example usage, output, and commentary for this tool 📚

🐧 OS 🐧

Linux 🐧

🧪 STABILITY 🧪

Unstable, still in development 🚧

🫶 AUTHORS 🫶

Brendan Gregg, Sasha Goldshtein, Rocky Xing 👑

🦋 SEE ALSO 🦋

hardirqs(8) 💅

---

bye bestie 💖 this was softirqs(8) from the USER COMMANDS section, dated 2015-10-20 🎀 slay forever ✨