Content-Type: multipart/alternative; boundary="_000_B4505C22A766DC4DBE0B2E038F983B6F035D0AE13C6Efs23airspan_" --_000_B4505C22A766DC4DBE0B2E038F983B6F035D0AE13C6Efs23airspan_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable I have a question about using oprofile (0.9.7) to track down unexpected pro= cess cpu usage. I am developing on the Cavium Octeon 4-core processor, with= one core running linux and 3 running simple executives. The problem is that a random linux process appears to be consuming signific= ant CPU resources, and using oprofile I can see that whatever activity is u= sing the resource is not visible to oprofile. The process does virtually no= thing except spend most of the time in a select() call awaiting some unix s= ocket activity. My question: Are there known holes to oprofile's monitoring? I suspect that= the Cavium Octeon mods to the kernel are the cause of this, making it look= like a process is hogging the CPU. Colin Wernham [cid:airspanlogo] Follow Airspan: [cid:facebook] [cid:twitter] This e-mail and any attachments may contain confidential and privileged inf= ormation and are intended solely for the intended recipient (s). If you are= not the intended recipient (s), please notify the sender immediately by re= turn e-mail, delete this e-mail and destroy any copies. Any dissemination o= r use of this information by a person other than the intended recipient (s)= is unauthorized and may be illegal. Airspan makes no representations or wa= rranties whatsoever about the content of this e-mail, its accuracy, complet= eness, security or freedom from viruses. --_000_B4505C22A766DC4DBE0B2E038F983B6F035D0AE13C6Efs23airspan_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

I have a questio= n about using oprofile (0.9.7) to track down unexpected process cpu usage. = I am developing on the Cavium Octeon 4-core processor, with one core runnin= g linux and 3 running simple executives.

 

The problem is that a random lin= ux process appears to be consuming significant CPU resources, and using opr= ofile I can see that whatever activity is using the resource is not visible= to oprofile. The process does virtually nothing except spend most of the t= ime in a select() call awaiting some unix socket activity.

 

My question: A= re there known holes to oprofile's monitoring? I suspect that the Cavium Oc= teon mods to the kernel are the cause of this, making it look like a proces= s is hogging the CPU.

 <= /p>

Colin Wernham


 

Follow Airspan:  =20

This e-mail and any attachments may contain confidential and privileged= =20 information and are intended solely for the intended recipient (s). If you = are=20 not the intended recipient (s), please notify the sender immediately by ret= urn=20 e-mail, delete this e-mail and destroy any copies. Any dissemination or use= of=20 this information by a person other than the intended recipient (s) is=20 unauthorized and may be illegal. Airspan makes no representations or warran= ties=20 whatsoever about the content of this e-mail, its accuracy, completeness,=20 security or freedom from viruses.


= --_000_B4505C22A766DC4DBE0B2E038F983B6F035D0AE13C6Efs23airspan_--