memory-leak-diagnosis

To diagnose memory leaks in Linux, this Skill analyzes memory metrics and allocation patterns across RSS, slab caches, and vmalloc. It uses a hypothesis-driven decision tree to validate specific leak scenarios and generate detailed root cause reports for both user-space and kernel space.

Gradual RSS growth often indicates memory leaks. This Skill automatically collects baseline metrics and isolates whether the root cause is kernel slab expansion, vmalloc leaks, cgroup memory pressure, or user-space heap allocation issues using system-level diagnostic tools.

The best way to identify kernel slab cache expansion or vmalloc leaks is through hypothesis-driven analysis. This Skill validates specific leak scenarios like heap fragmentation or circular references by analyzing allocation patterns to prevent system instability and OOM events.

Yes, you can analyze cgroup memory pressure and anonymous page leaks. This Skill supports multi-layer diagnosis across various memory metrics, detecting leaks within memory cgroups, anonymous pages, and other allocation layers to pinpoint the exact root cause.

To isolate kernel slab leaks versus user-space heap issues, provide the process PID to start the memory diagnosis. The Skill systematically collects memory metrics, applies a structured decision tree to validate hypotheses, and outputs a detailed root cause report identifying the specific leak source.