Discussion:
[Ipmitool-devel] 64 bit ARM support for ipmitool
Jim Mankovich
2014-02-17 22:13:27 UTC
Permalink
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
--
--- Jim Mankovich | ***@hp.com (US Mountain Time) ---
Zdenek Styblik
2014-02-18 19:12:25 UTC
Permalink
Post by Jim Mankovich
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
Jim,

I see no reason why it shouldn't be possible or why there should be
any problems. Having said that, I assume you're asking about compiler
options/cross-compilation?

Z.
Post by Jim Mankovich
--
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk
_______________________________________________
Ipmitool-devel mailing list
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
Jim Mankovich
2014-02-18 19:37:35 UTC
Permalink
Z,
I was simply trying to find out if anyone had done this or was currently
in the process of doing it.
You don't really need ipmitool on 64b ARM unless you have a 64b ARM
platform to test it on.
Post by Zdenek Styblik
Post by Jim Mankovich
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
Jim,
I see no reason why it shouldn't be possible or why there should be
any problems. Having said that, I assume you're asking about compiler
options/cross-compilation?
Z.
Post by Jim Mankovich
--
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk
_______________________________________________
Ipmitool-devel mailing list
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
--
--- Jim Mankovich | ***@hp.com (US Mountain Time) ---
Zdenek Styblik
2014-02-18 20:43:33 UTC
Permalink
Z,
I was simply trying to find out if anyone had done this or was currently in
the process of doing it.
You don't really need ipmitool on 64b ARM unless you have a 64b ARM platform
to test it on.
Debian is doing it as there are packages for arm64. I can't help you
more than that :)

Z.
Post by Zdenek Styblik
Post by Jim Mankovich
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
Jim,
I see no reason why it shouldn't be possible or why there should be
any problems. Having said that, I assume you're asking about compiler
options/cross-compilation?
Z.
Post by Jim Mankovich
--
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk
_______________________________________________
Ipmitool-devel mailing list
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
--
Jarrod B Johnson
2014-02-18 20:00:54 UTC
Permalink
Well, I will say that at least in one place, ipmitool assumes that the
system is little-endian. When iit goes to print GUID the GUID will come
out looking different whether the host is little or big endian. Of course
strictly speaking the IPMI format for DUID is not honored by anyone, but
the defacto behavior is the funky representation that DMI calls 'wire
format'. Based on the 'de facto' behavior ipmitool only looks right if
system is little endian.



From: Zdenek Styblik <***@gmail.com>
To: Jim Mankovich <***@hp.com>
Cc: ipmitool-devel <ipmitool-***@lists.sourceforge.net>
Date: 02/18/2014 02:13 PM
Subject: Re: [Ipmitool-devel] 64 bit ARM support for ipmitool
Post by Jim Mankovich
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
Jim,

I see no reason why it shouldn't be possible or why there should be
any problems. Having said that, I assume you're asking about compiler
options/cross-compilation?

Z.
Post by Jim Mankovich
--
------------------------------------------------------------------------------
Post by Jim Mankovich
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk
Post by Jim Mankovich
_______________________________________________
Ipmitool-devel mailing list
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
Rob Swindell
2014-02-18 20:59:00 UTC
Permalink
The big vs. little endian issue is a separate issue as ARM is bi-endian and as such an OS or application could be either (though not mixing endianess in one "machine"). Any big-endian issues with ipmitool should be observable on other existing big-endian platforms (e.g. PowerPC, MIPS, 68K, etc.).

Also, the GUID parsing/display issue is also a different issue as you mentioned ipmitool assumes a different field and byte order than that specified in Table 20-10 of the IPMI specification. Since existing BMCs seem to be tested primarily with ipmitool and the display output of the GUID appears consistent with other methods of querying/displaying the system GUID (e.g. via the BIOS setup or a BMC Web GUI), then the defacto behavior appears to be well entrenched even though it contradicts the spec. I agree, at the minimum, ipmitool should output the GUID consistently whether run on a big or little endian platform.

-Rob

From: Jarrod B Johnson [mailto:***@us.ibm.com]
Sent: Tuesday, February 18, 2014 12:01 PM
To: Zdenek Styblik
Cc: ipmitool-devel
Subject: Re: [Ipmitool-devel] 64 bit ARM support for ipmitool


Well, I will say that at least in one place, ipmitool assumes that the system is little-endian. When iit goes to print GUID the GUID will come out looking different whether the host is little or big endian. Of course strictly speaking the IPMI format for DUID is not honored by anyone, but the defacto behavior is the funky representation that DMI calls 'wire format'. Based on the 'de facto' behavior ipmitool only looks right if system is little endian.

[Inactive hide details for Zdenek Styblik ---02/18/2014 02:13:12 PM---On Mon, Feb 17, 2014 at 11:13 PM, Jim Mankovich <***@hp.]Zdenek Styblik ---02/18/2014 02:13:12 PM---On Mon, Feb 17, 2014 at 11:13 PM, Jim Mankovich <***@hp.com<mailto:***@hp.com>> wrote: > Has anyone looked into build

From: Zdenek Styblik <***@gmail.com<mailto:***@gmail.com>>
To: Jim Mankovich <***@hp.com<mailto:***@hp.com>>
Cc: ipmitool-devel <ipmitool-***@lists.sourceforge.net<mailto:ipmitool-***@lists.sourceforge.net>>
Date: 02/18/2014 02:13 PM
Subject: Re: [Ipmitool-devel] 64 bit ARM support for ipmitool

________________________________
Post by Jim Mankovich
Has anyone looked into building and testing ipmitool on any 64 bit arm
platform?
Jim,

I see no reason why it shouldn't be possible or why there should be
any problems. Having said that, I assume you're asking about compiler
options/cross-compilation?

Z.
Post by Jim Mankovich
--
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121054471&iu=/4140/ostg.clktrk
_______________________________________________
Ipmitool-devel mailing list
https://lists.sourceforge.net/lists/listinfo/ipmitool-devel
Loading...