Please wait for three weeks while this is reviewed.
Hank
On Mon, Apr 9, 2012 at 3:51 PM, Jim Mankovich <jm...@hp.com> wrote:
> All,
>
> Anyone care to code review this before I submit this to the CVS repository?
> I'm planning on submitting this on Wed April 11 unless I hear from someone
> before then.
>
> Prior Message,
>
> Attached is a patch for the TOB CVS repository to resolve how ipmitool
> incorrectly displays all Threshold
> Sensors as having analog readings instead of properly determining whether
> or not the Threshold Sensor
> is analog or discrete. This is a fairly large change and I would
> appreciate if I could get someone to
> do a code review for me. This code has already been reviewed by one
> other person here that is helping
> out with impitool bug fixes.
>
>
> In order to perform testing, I #ifdef'ed out the call to
> ipmi_sdr_get_sensor_reading
> in ipmi_sdr_get_sensor_reading_**ipmb and I plan on pushing this change
> to cvs as well unless someone
> has a better idea. This change makes the CVS code the same as 1.8.11
> again. I also ran across a
> defect ipmi_lanplus_send_payload() where timeout was getting reset to the
> default value on each invocation.
> I found a patch already existed for this, but the patch was flawed. I
> resolved this problem by loading
> the session timeout (after intf->open !) into a local variable instead of
> overwriting the session timeout in
> this function.
>
> This patch has been fairly well tested on a variety of HP Proliant
> Platforms and one Dell Platform.
> I would appreciate if others could try out this patch on other platforms
> and let me know if they see
> any problems. I used code coverage analysis in my testing to verify that
> the code changes I made
> were executed and I also verified that the ipmitool output for "sdr list",
> "sdr elist" and "sensor list"
> with -v, -c, no arguments, and -c with -v were correct when this patch is
> applied. The testing I did was
> mainly via the lanplus and lan interfaces.
>
> The following outlines the visible changes which will occur after
> application of this patch and
> some information pertaining to why the changes were necessary.
>
> Only Full Sensor Records of type Threshold can return Analog readings
>
> See Section 42.1 Event/Reading Type Codes
> Threshold == (SDR field - Event/Reading Type Code == 0x01)
> Changes to properly decode Threshold/Analog and Threshold/Discrete
>
>
> Full and Compact Sensor Records can both be of type Threshold
> Threshold can return Discrete or Analog readings
>
> See Sections 43.1 Full Sensor Records/43.2 Compact Sensor Records
> Discrete is identified by (SDR field - Sensor Units 1 [7:6] == 11b)
> Changes to properly decode Threshold/Discrete compact sensors
>
>
> Get Sensor Reading Command - Sensor reading should be ignored if the
> sensor does not return
> an numeric (analog) reading
>
> See Section 35.14 Get Sensor Reading Command
> "sdr list" prints "Sensor Reading" for Full Sensors (Discrete and
> Analog)
> Changes to display Sensor Reading for Compact
>
>
> verbose output OEM field is only printed for Discrete Sensors (Full and
> Compact)
>
> Changed to be printed for Threshold
>
>
> verbose output Event Message Control is only printed for Full Sensors
> (Discrete and Analog)
>
> Changed to be printed for Compact
>
>
> For Compact Sensors verbose output occurs instead of csv formatted when
> both -v and -c.
> For Full sensors, csv output is more verbose when you use -v with -c.
>
> Changed to make Compact Sensor output exactly the same as Full Sensor
> output when using -c with -v.
>
>
> Verbose output display for "Sensor Type" used the output string "Analog".
>
> Changed to correctly display "Sensor Type" as "Threshold" or "Discrete".
> "Analog" is no longer output as a "Sensor Type".
>
>
> Overview of the Code Changes
>
> The four separate sdr and sensor full and compact display routines have
> been merged into
> two separate routines. There is now a single routine for displaying a full
> and compact sdrs and
> a single routine for displaying full and compact sensors. These two
> routines were modified to
> properly decode analog and discrete Threshold sensors and Discrete
> sensors. The common
> header components of the compact and full SDR recodr structures were moved
> into a single new
> sdr record called sdr_record_common_sensor so that the common components
> of a compact and
> full sdr could be used by the new routines.
>
> --
> -- Jim Mankovich | jm...@hp.com --
>
>
>
> ------------------------------------------------------------------------------
> For Developers, A Lot Can Happen In A Second.
> Boundary is the first to Know...and Tell You.
> Monitor Your Applications in Ultra-Fine Resolution. Try it FREE!
> http://p.sf.net/sfu/Boundary-d2dvs2
>
> _______________________________________________
> Ipmitool-devel mailing list
> Ipmitool-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
>
>
------------------------------------------------------------------------------
Better than sec? Nothing is better than sec when it comes to
monitoring Big Data applications. Try Boundary one-second
resolution app monitoring today. Free.
http://p.sf.net/sfu/Boundary-dev2dev
_______________________________________________
Ipmitool-devel mailing list
Ipmitool-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
