[prev in list] [next in list] [prev in thread] [next in thread]
List: e1000-devel
Subject: Re: [E1000-devel] Technical query - driver version: kernel vs Intel
From: "Ronciak, John" <john.ronciak () intel ! com>
Date: 2010-04-23 19:49:00
Message-ID: DDC57477F5D6F845A0DDCB99D3C4812D0C9861DD60 () orsmsx510 ! amr ! corp ! intel ! com
[Download RAW message or body]
The versions of the drivers from support.intel.com and the Sourceforge versions are \
the same. The ones in an upstream kernel can be off slightly from what is on our SF \
and Intel sites due to timing of when changes get accepted into upstream kernels. To \
answer your question, use the SF drivers. They will run across a big range of kernel \
versions and will compile on just about all (if not all) distribution releases as \
well.
Cheers,
John
From: Spencer Goh [mailto:spencer.goh@ceptron.com]
Sent: Friday, April 23, 2010 11:43 AM
To: Ronciak, John
Cc: e1000-devel@lists.sourceforge.net
Subject: Re: [E1000-devel] Technical query - driver version: kernel vs Intel website \
& compile issues
Thanks v much for the quick reply.
So the s/f versions and the Intel versions... which is forked from which... and so \
which is more up to date... I'm guessing that the version numbers don't align? and \
so which is "better" for our 82576 cards? I know this prob depends on many \
factors... and do both drivers support the same feature sets / config parameter \
options... are they written completely independently of each other, or related code?
Cheers
Spencer
On 23 April 2010 19:38, Ronciak, John \
<john.ronciak@intel.com<mailto:john.ronciak@intel.com>> wrote: The problem with \
trying to update the ioatdma driver is that it is highly integrated with the OS \
version. We do not recommend doing what you are trying to do. The ioatdma driver is \
maintained completely in the upstream kernel since it so tied to that kernel. \
Stand-alone version of this driver don't work out very well. Please use the \
in-kernel version of it.
For our networking drivers like igb the Sourceforge and \
support.intel.com<http://support.intel.com> versions (stand-alone) are fully intended \
to be this way and you can easily update to the latest versions of them.
Cheers,
John
> -----Original Message-----
> From: Spencer Goh [mailto:spencer.goh@ceptron.com<mailto:spencer.goh@ceptron.com>]
> Sent: Friday, April 23, 2010 10:49 AM
> To: e1000-devel@lists.sourceforge.net<mailto:e1000-devel@lists.sourceforge.net>
> Subject: [E1000-devel] Technical query - driver version: kernel vs
> Intel website & compile issues
>
> Hi.
>
> I'm looking to get the best/latest drivers for our NICs which are using
> the Intel 82576 chipset.
> There appears to be a confusing versions when comparing to ones
> supplied in kernel as part of our Fedora 12 OS.
> NB - I raised the version discrepancy with via Intel Tech Support chat
> they suggested I contact thsi list to give you more detail on the
> issue.
>
> Fedora 12 - 2.6.32.11-99.fc12.x86_64
> ioatdma - 4.00 (newer?)
> igb - 1.3.16-k2 (older?)
>
> Intel Website - source code
> ioatdma 3.83.0 -
> http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=10959&Pr
> odId=3024&lang=eng
> igb 2.1.9 -
> http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=13663&Pr
> odId=3024&lang=eng
>
> Compile source downloaded from website:
>
> IGB compiles fine from source: OK
> IOATDMA fails to compile. Error show below:
>
> If I compile IGB not against ioatdma 3.83, but stock kernel driver
> version 4.00 is this ok, will it function correctly? Why is ioatdma
> newer than Intel published src?
> How to proceed with ioatdma module? try further to compile 3.83 or use
> 4.00 confused....
>
> Thanks
> Spencer
>
> ============
> [153]$ lspci -v | grep 82576 -A 16
> 01:00.0 Ethernet controller: Intel Corporation 82576 Gigabit Network
> Connection (rev 01)
> Subsystem: Super Micro Computer Inc Device 0100
> Flags: bus master, fast devsel, latency 0, IRQ 16
> Memory at f8ee0000 (32-bit, non-prefetchable) [size=128K]
> Memory at f8ec0000 (32-bit, non-prefetchable) [size=128K]
> I/O ports at dc00 [size=32]
> Memory at f8ebc000 (32-bit, non-prefetchable) [size=16K]
> Expansion ROM at f8e80000 [disabled] [size=128K]
> Capabilities: [40] Power Management version 3
> Capabilities: [50] MSI: Enable- Count=1/1 Maskable+ 64bit+
> Capabilities: [70] MSI-X: Enable+ Count=10 Masked-
> Capabilities: [a0] Express Endpoint, MSI 00
> Capabilities: [100] Advanced Error Reporting
> Capabilities: [140] Device Serial Number XXXXXXXXXXXXXXXXXXXX
> Capabilities: [150] Alternative Routing-ID Interpretation (ARI)
> Capabilities: [160] Single Root I/O Virtualization (SR-IOV)
> Kernel driver in use: igb
>
> [119] /usr/src/ioatdma-3.83 $ make
> make -C dca configure
> make[1]: Nothing to be done for `configure'.
> make -C ioatdma configure
> make[1]: Nothing to be done for `configure'.
> make -C /lib/modules/2.6.32.11-99.fc12.x86_64/source
> M=/usr/src/ioatdma-3.83 modules
> CC [M] /usr/src/ioatdma-3.83/ioatdma/ioat_dma.o
> /usr/src/ioatdma-3.83/ioatdma/ioat_dma.c: In function
> 'ioat_dma_alloc_descriptor':
> /usr/src/ioatdma-3.83/ioatdma/ioat_dma.c:1006: error: 'struct
> dma_async_tx_descriptor' has no member named 'tx_list'
> make[3]: *** [/usr/src/ioatdma-3.83/ioatdma/ioat_dma.o] Error 1
> make[2]: *** [/usr/src/ioatdma-3.83/ioatdma] Error 2
> make[1]: *** [_module_/usr/src/ioatdma-3.83] Error 2
> make: *** [modules] Error 2
>
>
> [137] /usr/src/igb-2.1.9/src $ make
> make -C /lib/modules/2.6.32.11-99.fc12.x86_64/build
> SUBDIRS=/usr/src/igb-2.1.9/src modules
> make[1]: Entering directory `/usr/src/kernels/2.6.32.11-99.fc12.x86_64'
> CC [M] /usr/src/igb-2.1.9/src/igb_main.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_82575.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_mac.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_nvm.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_phy.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_manage.o
> CC [M] /usr/src/igb-2.1.9/src/igb_param.o
> CC [M] /usr/src/igb-2.1.9/src/igb_ethtool.o
> CC [M] /usr/src/igb-2.1.9/src/kcompat.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_api.o
> CC [M] /usr/src/igb-2.1.9/src/e1000_mbx.o
> LD [M] /usr/src/igb-2.1.9/src/igb.o
> Building modules, stage 2.
> MODPOST 1 modules
> CC /usr/src/igb-2.1.9/src/igb.mod.o
> LD [M] /usr/src/igb-2.1.9/src/igb.ko
> make[1]: Leaving directory `/usr/src/kernels/2.6.32.11-99.fc12.x86_64'
>
>
>
> [140] /usr/src/igb-2.1.9/src $ make CFLAGS_EXTRA="-DIGB_DCA"
> CHK include/linux/version.h
> CHK include/linux/utsrelease.h
> SYMLINK include/asm -> include/asm-x86
> make[1]: *** No rule to make target `missing-syscalls'. Stop.
> make: *** [prepare0] Error 2
>
>
> =======
> Jorge: I have created a case for the chat it is 9683565
> Jorge: I will have to verify with engineering your question, you can
> also try sending an e-mail to igb-devel@lists.sf.net<mailto:igb-devel@lists.sf.net> \
> =======
> [119]$ make
> make -C dca configure
> make[1]: Nothing to be done for `configure'.
> make -C ioatdma configure
> make[1]: Nothing to be done for `configure'.
> make -C /lib/modules/2.6.32.11-99.fc12.x86_64/source
> M=/usr/src/ioatdma-3.83 modules
> CC [M] /usr/src/ioatdma-3.83/ioatdma/ioat_dma.o
> /usr/src/ioatdma-3.83/ioatdma/ioat_dma.c: In function
> 'ioat_dma_alloc_descriptor':
> /usr/src/ioatdma-3.83/ioatdma/ioat_dma.c:1006: error: 'struct
> dma_async_tx_descriptor' has no member named 'tx_list'
> make[3]: *** [/usr/src/ioatdma-3.83/ioatdma/ioat_dma.o] Error 1
> make[2]: *** [/usr/src/ioatdma-3.83/ioatdma] Error 2
> make[1]: *** [_module_/usr/src/ioatdma-3.83] Error 2
> make: *** [modules] Error 2
>
>
> [120]$ uname -a
> Linux 2.6.32.11-99.fc12.x86_64 #1 SMP Mon Apr 5 19:59:38 UTC 2010
> x86_64
> x86_64 x86_64 GNU/Linux
>
>
> [121]$ modinfo igb
> filename:
> /lib/modules/2.6.32.11-99.fc12.x86_64/kernel/drivers/net/igb/igb.ko
> version: 1.3.16-k2
> license: GPL
> description: Intel(R) Gigabit Ethernet Network Driver
> author: Intel Corporation, \
> <e1000-devel@lists.sourceforge.net<mailto:e1000-devel@lists.sourceforge.net>>
>
> srcversion: 4BFEA86B90B943014153402
> alias: pci:v00008086d000010D6sv*sd*bc*sc*i*
> alias: pci:v00008086d000010A9sv*sd*bc*sc*i*
> alias: pci:v00008086d000010A7sv*sd*bc*sc*i*
> alias: pci:v00008086d000010E8sv*sd*bc*sc*i*
> alias: pci:v00008086d0000150Dsv*sd*bc*sc*i*
> alias: pci:v00008086d000010E7sv*sd*bc*sc*i*
> alias: pci:v00008086d000010E6sv*sd*bc*sc*i*
> alias: pci:v00008086d0000150Asv*sd*bc*sc*i*
> alias: pci:v00008086d000010C9sv*sd*bc*sc*i*
> depends: dca
> vermagic: 2.6.32.11-99.fc12.x86_64 SMP mod_unload
> parm: max_vfs:Maximum number of virtual functions to allocate
> per
> physical function (uint)
>
> [122]$ modinfo ioatdma
> filename:
> /lib/modules/2.6.32.11-
> 99.fc12.x86_64/kernel/drivers/dma/ioat/ioatdma.ko
> author: Intel Corporation
> license: Dual BSD/GPL
> version: 4.00
> srcversion: 7380D4CDD80EC629A965621
> alias: pci:v00008086d00003719sv*sd*bc*sc*i*
> alias: pci:v00008086d00003718sv*sd*bc*sc*i*
> alias: pci:v00008086d00003717sv*sd*bc*sc*i*
> alias: pci:v00008086d00003716sv*sd*bc*sc*i*
> alias: pci:v00008086d00003715sv*sd*bc*sc*i*
> alias: pci:v00008086d00003714sv*sd*bc*sc*i*
> alias: pci:v00008086d00003713sv*sd*bc*sc*i*
> alias: pci:v00008086d00003712sv*sd*bc*sc*i*
> alias: pci:v00008086d00003711sv*sd*bc*sc*i*
> alias: pci:v00008086d00003710sv*sd*bc*sc*i*
> alias: pci:v00008086d0000342Csv*sd*bc*sc*i*
> alias: pci:v00008086d0000342Bsv*sd*bc*sc*i*
> alias: pci:v00008086d0000342Asv*sd*bc*sc*i*
> alias: pci:v00008086d00003429sv*sd*bc*sc*i*
> alias: pci:v00008086d00003433sv*sd*bc*sc*i*
> alias: pci:v00008086d00003432sv*sd*bc*sc*i*
> alias: pci:v00008086d00003431sv*sd*bc*sc*i*
> alias: pci:v00008086d00003430sv*sd*bc*sc*i*
> alias: pci:v00008086d0000402Fsv*sd*bc*sc*i*
> alias: pci:v00001018d0000001Csv*sd*bc*sc*i*
> alias: pci:v00008086d000065FFsv*sd*bc*sc*i*
> alias: pci:v00008086d0000360Bsv*sd*bc*sc*i*
> alias: pci:v00008086d00001A38sv*sd*bc*sc*i*
> depends: dca
> vermagic: 2.6.32.11-99.fc12.x86_64 SMP mod_unload
> parm: ioat_ring_alloc_order:ioat2+: allocate 2^n descriptors
> per
> channel (default: 8 max: 16) (int)
> parm: ioat_ring_max_alloc_order:ioat2+: upper limit for ring
> size
> (default: 16) (int)
> parm: ioat_pending_level:high-water mark for pushing ioat
> descriptors (default: 4) (int)
> parm: ioat_initial_desc_count:ioat1: initial descriptors per
> channel (default: 256) (int)
> parm: ioat_interrupt_style:set ioat interrupt style: msix
> (default), msix-single-vector, msi, intx) (string)
> parm: ioat_dca_enabled:control support of dca service
> (default: 1)
> (int)
[Attachment #3 (--===============1864859292138571527==)]
------------------------------------------------------------------------------
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic