From 04490a99946e81b22dd7c512de8157cfd91501a6 Mon Sep 17 00:00:00 2001 From: Vivek Ankit <72314052+vian9@users.noreply.github.com> Date: Sun, 27 Jun 2021 04:15:57 +0530 Subject: [PATCH] Improved Grammatical Errors --- README | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/README b/README index 3163e4e..02b481d 100644 --- a/README +++ b/README @@ -33,14 +33,14 @@ syscall introduced in a mature form into the Linux Kernel mainline in 2009. This interface was previously only usable via a binary tool called 'perf' included in Linux Kernel source distributions today. -There are several key shortcomings of interfacing with the system call only -through a controlled binary. The first is that the granularity of tracing +There are several key shortcoming of interfacing with the system call only +through a controlled binary. The first is that the granularity of tracing hardware performance counter registers and software tracepoints is at a whole -binary level. There is no method for tracing portions of code within the -application if desired. This is because there is no nice API exposed to +binary level. There is no method for tracing portions of code within the +application if desired. This is because there is no nice API exposed to userspace developers that can interface with this system call. The second shortcoming is that the tool, although extensive, may not account for every -possible use case for these counters and tracepoints. Offering a +possible use case for these counters and tracepoints. Offering a non-restrictive API to userspace gives developers full power and freedom to implement any functionality desired.