Rafi, Albert
I have to run now. But that dump tells us that the mmap reads are
returning garbage.
I assume you are running this clean out of power up and NOT after any
failed attempt of running loadBitstream?
If that is true; it is as if you somehow have a defective *default*
bitstream. (Where did it come from?)
I may not be able to debug this with you side by side until I'm back
in my lab Thursday.
Everything looks fine; except for the garbage coming back.
Adding the list back so others may offer a debug suggestion.
-Shep
On Mon, Jun 13, 2011 at 12:09 PM, Rafi Rubin <[email protected]
<mailto:[email protected]>> wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 06/13/11 12:01, Albert Kwon wrote:
> Hi Rafi,
>
> Can you run the following commands as sudo for me? Shep
wants me to
> test this out by running things actually as sudo, not what we've
been doing.
>
> Environment variable that needs to be set:
> export
> XIL_PAR_DESIGN_CHECK_VERBOSE=1
> export XIL_TIMING_ALLOW_IMPOSSIBLE=1
> export PATH=$PATH:/scratch/ic3/ocpi/bin/
>
> Then inside /scratch/ic3/ocpi/script, run
> sudo -E ./ocfrp_check,
Error during probe for OCFRP: Magic numbers do not match in
region/bar 0
Did not find any OpenCPI FPGA reference platform boards.
> sudo -E *swctl **0xFA000000 0xFB9F0000 admin*
OCCP Admin Space
Open: 0xffffffff "ÿÿÿÿ"
CPI: 0xffffffff "ÿÿÿÿ"
revision: 0xffffffff
birthday: 0xffffffff Sun Feb 7 06:28:15 2106
workerMask: 0xffffffff workers 0 1 2 3 4 5 6 7 8 9 10 11 12 13
14 15
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 exist
pci_dev_id: 0xffffffff
attention: 0xffffffff
cpStatus: 0xffffffff
scratch20: 0xffffffff
scratch24: 0xffffffff
cpControl: 0xffffffff
timeStatus: 0xffffffff ppsLostSticky gpsInSticky ppsInSticky
timeSetSticky ppsOK ppsLost
timeControl: 0xffffffff
gpsTimeMS: 0xffffffff (4294967295) Sun Feb 7 06:28:15 2106
gpsTimeLS: 0xffffffff (4294967295)
deltaTimeMS: 0xffffffff
deltaTimeLS: 0xffffffff
refPerPPS: 0xffffffff (-1)
numDPMemReg: 0xffffffff (-1)
UUID[ 3: 0]: 0xffffffff 0xffffffff 0xffffffff 0xffffffff
UUID[ 7: 4]: 0xffffffff 0xffffffff 0xffffffff 0xffffffff
UUID[11: 8]: 0xffffffff 0xffffffff 0xffffffff 0xffffffff
UUID[15:12]: 0xffffffff 0xffffffff 0xffffffff 0xffffffff
> sudo lspci -v
$ sudo lspci -vs 05:00.0
/scratch/ic3/ocpi/scripts 12:06PM hex
05:00.0 RAM memory: Xilinx Corporation Device 4243 (rev 02)
Subsystem: Xilinx Corporation Device 0007
Flags: fast devsel, IRQ 10
[virtual] Memory at fa000000 (32-bit, non-prefetchable)
[size=16M]
[virtual] Memory at fb9f0000 (32-bit, non-prefetchable)
[size=64K]
Capabilities: [40] Power Management version 3
Capabilities: [48] MSI: Enable- Count=1/1 Maskable- 64bit+
Capabilities: [60] Express Endpoint, MSI 01
Capabilities: [100] Device Serial Number
00-00-00-01-01-00-0a-35
$ sudo lspci -vvs 05:00.0
/scratch/ic3/ocpi/scripts 12:06PM hex
05:00.0 RAM memory: Xilinx Corporation Device 4243 (rev 02)
Subsystem: Xilinx Corporation Device 0007
Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
Interrupt: pin A routed to IRQ 10
Region 0: [virtual] Memory at fa000000 (32-bit,
non-prefetchable) [size=16M]
Region 1: [virtual] Memory at fb9f0000 (32-bit,
non-prefetchable) [size=64K]
Capabilities: [40] Power Management version 3
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA
PME(D0+,D1+,D2+,D3hot+,D3cold-)
Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [48] MSI: Enable- Count=1/1 Maskable- 64bit+
Address: 0000000000000000 Data: 0000
Capabilities: [60] Express (v2) Endpoint, MSI 01
DevCap: MaxPayload 512 bytes, PhantFunc 0, Latency L0s
<64ns, L1 unlimited
ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset-
DevCtl: Report errors: Correctable- Non-Fatal- Fatal-
Unsupported-
RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
MaxPayload 128 bytes, MaxReadReq 512 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq-
AuxPwr-
TransPend-
LnkCap: Port #0, Speed 5GT/s, Width x4, ASPM L0s,
Latency L0 unlimited, L1 unlimited
ClockPM- Surprise- LLActRep- BwNot-
LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- Retrain-
CommClk-
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 5GT/s, Width x4, TrErr- Train- SlotClk-
DLActive- BWMgmt- ABWMgmt-
DevCap2: Completion Timeout: Range B, TimeoutDis-
DevCtl2: Completion Timeout: 50us to 50ms, TimeoutDis-
LnkCtl2: Target Link Speed: 5GT/s, EnterCompliance-
SpeedDis-, Selectable De-emphasis: -6dB
Transmit Margin: Normal Operating Range,
EnterModifiedCompliance- ComplianceSOS-
Compliance De-emphasis: -6dB
LnkSta2: Current De-emphasis Level: -6dB
Capabilities: [100 v1] Device Serial Number
00-00-00-01-01-00-0a-35
>
> and if you can send me the outputs of each, that would be great
> (only the output of Xilinx for lspci -v).
>
> And after the three commands above, if you don't mind,
please try to run
> sudo ./loadBitStreamML605_sudo
> ../build/ml605-20110609_1313/fpgaTop.bit 0000:05:00.0 auto
> inside the same script folder.
>
> Thanks a lot.
> Albert
>
>
> -------- Original Message --------
> Subject: Re: OCPI Question
> Date: Mon, 13 Jun 2011 11:52:18 -0400
> From: Shepard Siegel <[email protected]
<mailto:[email protected]>>
> To: Albert Kwon <[email protected]
<mailto:[email protected]>>
>
>
>
> right. good
> so (with Rafi's help) try from $OCPI_DIR/bin
>
> *$ sudo -E ./swctl 0xFA000000 0xFB9F0000 admin*
>
> This should dump out a bunch of registers with sensible values,
similar
> to this
>
> [shep@core960 bin]$ sudo ./swctl 0xF1000000 0xF2FF0000 admin
>
> OCCP Admin Space
> Open: 0x4f70656e "nepO"
> CPI: 0x43504900 ""
> revision: 0x00000001
> birthday: 0x4b2d41ab Sat Dec 19 21:12:11 2009
> workerMask: 0x0000601f workers 0 1 2 3 4 13 14 exist
> pci_dev_id: 0x00000300
> attention: 0x00000000
> cpStatus: 0x00000381
> scratch20: 0x00000000
> scratch24: 0x00000000
> cpControl: 0x2202f798
> timeStatus: 0x00000000
> timeControl: 0x00000000
> gpsTimeMS: 0x0000fd6c (64876) Thu Jan 1 18:01:16 1970
> gpsTimeLS: 0x06ba5820 (112874480)
> deltaTimeMS: 0x00000000
> deltaTimeLS: 0x00000000
> refPerPPS: 0x4b4a4948 (1263159624)
> numDPMemReg: 0xdeadc0de (-559038242)
> UUID[ 3: 0]: 0x4f4e4d4c 0x4b4a4948 0x00000000 0x00000000
> UUID[ 7: 4]: 0xdeadc0de 0xdeadc0de 0xdeadc0de 0xdeadc0de
> UUID[11: 8]: 0xdeadc0de 0xdeadc0de 0xdeadc0de 0xdeadc0de
> UUID[15:12]: 0xdeadc0de 0xdeadc0de 0xdeadc0de 0xdeadc0de
>
>
>
>
> On Mon, Jun 13, 2011 at 11:47 AM, Albert Kwon
<[email protected] <mailto:[email protected]>
> <mailto:[email protected] <mailto:[email protected]>>>
wrote:
>
> I will ask Rafi to run those things in sudo, and see what
happens.
> This is the output of lspci -v:
>
>
> 05:00.0 RAM memory: Xilinx Corporation Device 4243 (rev 02)
> Subsystem: Xilinx Corporation Device 0007
> Flags: fast devsel, IRQ 10
> [virtual] Memory at fa000000 (32-bit, non-prefetchable)
[size=16M]
> [virtual] Memory at fb9f0000 (32-bit, non-prefetchable)
[size=64K]
> Capabilities: <access denied>
>
> It seems like to get the full output, I need to run this as
sudo as
> well. Also, are the bar0 and bar1 in this case 0xFA000000 and
> 0xFB9F0000 respectively?
>
> Thanks,
> Albert
>
>
>
> On 06/13/2011 11:32 AM, Shepard Siegel wrote:
>> Albert,
>> It's highly unusual for lspci to see the board just fine; and
>> ocfrp_check not find it. Especially with the default bitstream.
>> If you can, do it as root with sudo -E, not Rafi's alternative.
>> Also, can you post the output from lspci -v (but just for the
>> Xlinix device)
>> There is a compiled version of swctl in the $OCPI_DIR/bin
directory
>> Or you can go the /src dir, say $gcc swctl.c; mv a.out
../bin/swctl
>> In any event, you will need to know the BAR0 and BAR1
addresses as
>> reported by lspci -v and say
>>
>> $ sudo -E ./swctl 0xDE000000 0xDFFF0000 admin
>>
>> (I just made up the Bar addresses, you must use the correct
ones)
>>
>> And post the output you get here with the default flashed
bitstream.
>>
>> Thanks.
>> _Shep
>>
>>
>> On Mon, Jun 13, 2011 at 11:22 AM, Albert Kwon
>> <[email protected] <mailto:[email protected]>
<mailto:[email protected] <mailto:[email protected]>>>
wrote:
>>
>> ocfrp_check gives the same result:
>>
>>
>> Error during probe for OCFRP: Magic numbers do not match in
>> region/bar 0
>> Did not find any OpenCPI FPGA reference platform boards.
>>
>> And I can't find the command/script swctl. Do I need to do
>> something to get this program/script?
>>
>> Thanks so much for the help.
>> -Albert
>>
>>
>> On 06/13/2011 11:14 AM, Shepard Siegel wrote:
>>> Albert,
>>> I've never done it that way but Rafi must know what
he is doing.
>>>
>>> Let's back up one more time before we move ahead.
Before we
>>> try to load your own bitstream with loadBitstreamML605, do
>>> these programs work as expected on the default flash
bitstream:
>>>
>>> ocfrp_check (to see the board)
>>> swctl <bar0> <bar1> admin
>>>
>>> If not, then before we debug loadBitstreamXXX, we
should get
>>> ocfrp_check and swctl working.
>>>
>>> -Shep
>>>
>>> note: sharing with opencpi_dev for others to chime in.
>>>
>>>
>>>
>>>
>>> On Mon, Jun 13, 2011 at 10:59 AM, Albert Kwon
>>> <[email protected] <mailto:[email protected]>
<mailto:[email protected] <mailto:[email protected]>>>
wrote:
>>>
>>> Instead of having to run these things as sudo
(since it's
>>> Rafi's machine that we are running this on), Rafi
set the
>>> suid on ocfrp_check, and put it under
/opt/opencpi/bin.
>>> So on the script, instead of doing sudo ./ocfrp_check
>>> $device, I do /opt/opencpi/bit/ocfrp_check
$device. I've
>>> also asked him to change permission on
>>> /sys/bus/pci/devices/$device/config, so that I can
write
>>> to it. I've also changed the location of $tmp for
>>> loadBitStreamML605 script to my home folder's tmp. So
>>> currently, I don't think there are any commands that
>>> require sudo.
>>>
>>> environment OCPI_DIR is currently set to be
>>> /scratch/ic3/ocpi.
>>>
>>> Is there something wrong with that we are doing here?
>>>
>>> -Albert
>>>
>>>
>>>
>>> On Sat, Jun 11, 2011 at 7:58 PM, Albert Kwon
>>> <[email protected] <mailto:[email protected]>
<mailto:[email protected] <mailto:[email protected]>>>
wrote:
>>>
>>> Hi Shep,
>>>
>>> The bitstream upload failed once due to some sudo
>>> privilege issues, and now when I run loadBitStream
>>> script, I get this error:
>>>
>>> Checking existing bitstream..
>>> Error during probe for OCFRP: Magic numbers do not
match
>>> in region/bar 0
>>> Did not find an OpenCPI FPGA reference platform board
>>> named "0000:05:00.0".
>>> Existing bitstream does not appear to be functional.
>>> There is no config snapshot in
>>> /opt/cpi/ocfrp/pci-config-snapshots/0000:05:00.0
to restore.
>>>
>>> Do you have any suggestion as to what to try?
>>>
>>> Thanks,
>>> Albert
>>>
>>>
>>>
>>> snip
>>
>>
>
>
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/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=sy9Z
-----END PGP SIGNATURE-----
