From df5902545c718f3a92f6423aae1f43d8144bd53c Mon Sep 17 00:00:00 2001 From: Andrey Konovalov Date: Fri, 12 Jun 2020 14:48:32 +0200 Subject: Update syscall_descriptions.md --- docs/syscall_descriptions.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'docs') diff --git a/docs/syscall_descriptions.md b/docs/syscall_descriptions.md index 78e46e832..77abbad79 100644 --- a/docs/syscall_descriptions.md +++ b/docs/syscall_descriptions.md @@ -57,11 +57,12 @@ To enable fuzzing of a new kernel interface: 1. Study the interface, find out which syscalls are required to use it. Sometimes there is nothing besides the source code, but here are some things that may help: - - Searching the Internet for the subsystem name and/or some unique constants. + - Searching the Internet for the interface name and/or some unique constants. - Grepping Documentation/ dir in the kernel. - Searching tools/testing/ dir in the kernel. - Looking for large comment blocks in the source code. - Finding commit that added the interface via `git blame` or `git log` and reading the commit description. + - Reading source code of or tracing libraries or applications that are known to use this interface. 2. Using [syntax documentation](syscall_descriptions_syntax.md) and [existing descriptions](/sys/linux/) as an example, add a declarative -- cgit mrf-deployment