Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

NEHALEM arch detection #69

Closed
ViralBShah opened this issue Oct 28, 2011 · 19 comments
Closed

NEHALEM arch detection #69

ViralBShah opened this issue Oct 28, 2011 · 19 comments
Assignees

Comments

@ViralBShah
Copy link
Contributor

The build process seems to be unable to detect NEHALEM correctly. I have found issues on a recent macbook air (core i7) and on an ubuntu 11.0

make -C openblas-v0.1alpha2.2 USE_THREAD=0 NO_LAPACK=1 CC=gcc FC=gfortran
nm: '/lib/libpthread.so_': No such file
getarch_2nd.c: In function ‘main’:
getarch_2nd.c:12:35: error: ‘SGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:12:35: note: each undeclared identifier is reported only once for each function it appears in
getarch_2nd.c:13:35: error: ‘SGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:14:35: error: ‘DGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:15:35: error: ‘DGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:19:35: error: ‘CGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:20:35: error: ‘CGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:21:35: error: ‘ZGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:22:35: error: ‘ZGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:29:50: error: ‘SGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:30:50: error: ‘DGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:31:50: error: ‘CGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:32:50: error: ‘ZGEMM_DEFAULT_Q’ undeclared (first use in this function)
make[1]: *_* [getarch_2nd] Error 1
make[1]: Entering directory /home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2' Makefile:107: *** OpenBLAS: Detecting CPU failed. Please set TARGET explicitly, e.g. make TARGET=your_cpu_target. Please read README for the detail.. Stop. make[1]: Leaving directory/home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2'
make: *** [openblas-v0.1alpha2.2/libopenblas.a] Error 2

@xianyi
Copy link
Collaborator

xianyi commented Oct 29, 2011

Hi Viral,

What's the version? Is it v0.1alpha2.2? Could you try the lastest
version or develop branch on github.com?

Thank you for the report.

Xianyi

Viral B. Shah write:Sent: 2011/10/28 23:53

The build process seems to be unable to detect NEHALEM correctly. I have found issues on a recent macbook air (core i7) and on an ubuntu 11.0

make -C openblas-v0.1alpha2.2 USE_THREAD=0 NO_LAPACK=1 CC=gcc FC=gfortran
nm: '/lib/libpthread.so_': No such file
getarch_2nd.c: In function ‘main’:
getarch_2nd.c:12:35: error: ‘SGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:12:35: note: each undeclared identifier is reported only once for each function it appears in
getarch_2nd.c:13:35: error: ‘SGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:14:35: error: ‘DGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:15:35: error: ‘DGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:19:35: error: ‘CGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:20:35: error: ‘CGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:21:35: error: ‘ZGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:22:35: error: ‘ZGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:29:50: error: ‘SGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:30:50: error: ‘DGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:31:50: error: ‘CGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:32:50: error: ‘ZGEMM_DEFAULT_Q’ undeclared (first use in this function)
make[1]: *_* [getarch_2nd] Error 1
make[1]: Entering directory /home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2' Makefile:107: *** OpenBLAS: Detecting CPU failed. Please set TARGET explicitly, e.g. make TARGET=your_cpu_target. Please read README for the detail.. Stop. make[1]: Leaving directory/home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2'
make: *** [openblas-v0.1alpha2.2/libopenblas.a] Error 2

@ViralBShah
Copy link
Contributor Author

Yes we were on 0.1alpha2.2. I just updated to 2.4, and am waiting to hear back from some people who are trying it out, as I don't have a nehalem box myself.

-viral

On Oct 29, 2011, at 6:06 AM, Xianyi Zhang wrote:

Hi Viral,

What's the version? Is it v0.1alpha2.2? Could you try the lastest
version or develop branch on github.com?

Thank you for the report.

Xianyi

Viral B. Shah write:Sent: 2011/10/28 23:53

The build process seems to be unable to detect NEHALEM correctly. I have found issues on a recent macbook air (core i7) and on an ubuntu 11.0

make -C openblas-v0.1alpha2.2 USE_THREAD=0 NO_LAPACK=1 CC=gcc FC=gfortran
nm: '/lib/libpthread.so_': No such file
getarch_2nd.c: In function ‘main’:
getarch_2nd.c:12:35: error: ‘SGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:12:35: note: each undeclared identifier is reported only once for each function it appears in
getarch_2nd.c:13:35: error: ‘SGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:14:35: error: ‘DGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:15:35: error: ‘DGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:19:35: error: ‘CGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:20:35: error: ‘CGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:21:35: error: ‘ZGEMM_DEFAULT_UNROLL_M’ undeclared (first use in this function)
getarch_2nd.c:22:35: error: ‘ZGEMM_DEFAULT_UNROLL_N’ undeclared (first use in this function)
getarch_2nd.c:29:50: error: ‘SGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:30:50: error: ‘DGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:31:50: error: ‘CGEMM_DEFAULT_Q’ undeclared (first use in this function)
getarch_2nd.c:32:50: error: ‘ZGEMM_DEFAULT_Q’ undeclared (first use in this function)
make[1]: *_* [getarch_2nd] Error 1
make[1]: Entering directory /home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2' Makefile:107: *** OpenBLAS: Detecting CPU failed. Please set TARGET explicitly, e.g. make TARGET=your_cpu_target. Please read README for the detail.. Stop. make[1]: Leaving directory/home/stephan/Desktop/julia/external/openblas-v0.1alpha2.2'
make: *** [openblas-v0.1alpha2.2/libopenblas.a] Error 2

Reply to this email directly or view it on GitHub:
#69 (comment)

@xianyi
Copy link
Collaborator

xianyi commented Nov 4, 2011

Hi,

Do you or your colleague meet the problem on Nehalem again?

Thanks
Xianyi

@ViralBShah
Copy link
Contributor Author

I built with DYNAMIC_ARCH=1, and things don't crash anymore. I assume it is using the right version - but don't have a way to tell.

-viral

On Nov 4, 2011, at 12:56 PM, Xianyi Zhang wrote:

Hi,

Do you or your colleague meet the problem on Nehalem again?

Thanks
Xianyi


Reply to this email directly or view it on GitHub:
#69 (comment)

@ViralBShah
Copy link
Contributor Author

BTW, julia users still keep reporting this error. If using DYNAMIC_ARCH=1, is it possible that this can be made smooth by detecting at runtime, and using a safe fallback in case the detection fails, while printing a warning?

@xianyi
Copy link
Collaborator

xianyi commented Jan 26, 2012

Hi Viral,

Thank you for this report.
Now we are in Spring Festival holiday. I will check this bug in next month.

Thank you
Xianyi

在 2012-1-26,11:16,"Viral B. Shah"reply@reply.github.com 写道:

BTW, julia users still keep reporting this error. If using DYNAMIC_ARCH=1, is it possible that this can be made smooth by detecting at runtime, and using a safe fallback in case the detection fails, while printing a warning?


Reply to this email directly or view it on GitHub:
#69 (comment)

@ViralBShah
Copy link
Contributor Author

Ok, I now have a NEHALEM machine and am using 0.1alpha2.4, and it fails with the same error.

@xianyi
Copy link
Collaborator

xianyi commented Feb 13, 2012

Hi Viral,

I think Mac Air 2011 equips with Intel Core i5-2557M (http://ark.intel.com/products/54620). This CPU is Sandy Bridge microarchitecture.

Because we didn't optimize OpenBLAS on Sandy bridge. it couldn't automatically detect this CPU.

Thus, you should set "TARGET" flag on compiling, e.g.

make TARGET=NEHALEM

Xianyi

@ViralBShah
Copy link
Contributor Author

It would be nice if NEHALEM is set automatically for Sand bridge arch,
with a warning message. Currently, this just breaks the build, and it
would be nice if that could be avoided. Would that be possible?

Thanks,

-viral

On Mon, Feb 13, 2012 at 2:38 PM, Xianyi Zhang
reply@reply.github.com
wrote:

Hi Viral,

I think Mac Air 2011 equips with Intel Core i5-2557M (http://ark.intel.com/products/54620). This CPU is Sandy Bridge microarchitecture.

Because we didn't optimize OpenBLAS on Sandy bridge. it couldn't automatically detect this CPU.

Thus, you should set "TARGET" flag on compiling, e.g.

make TARGET=NEHALEM

Xianyi


Reply to this email directly or view it on GitHub:
#69 (comment)

-viral

@xianyi
Copy link
Collaborator

xianyi commented Feb 13, 2012

Hi Viral,

OK. Pleas give me your CPU processor number, e.g. i5-2557M.
I will add detecting this cpuid into OpenBLAS.

Thanks

Xianyi

@ViralBShah
Copy link
Contributor Author

It has also been reported by people using Linux etc. Can we make it
default to NEHALEM for all CPUs later than Nehalem, until support is
added?

-viral

On Mon, Feb 13, 2012 at 2:46 PM, Xianyi Zhang
reply@reply.github.com
wrote:

Hi Viral,

OK. Pleas give me your CPU processor number, e.g. i5-2557M.
I will add detecting this cpuid into OpenBLAS.

Thanks

Xianyi


Reply to this email directly or view it on GitHub:
#69 (comment)

-viral

@ghost ghost assigned xianyi Feb 13, 2012
xianyi added a commit that referenced this issue Feb 13, 2012
@xianyi
Copy link
Collaborator

xianyi commented Feb 13, 2012

Hi Viral,

I just added those CPUs. You can download it on develop branch.
Could you help me test it?

Thanks
Xianyi

@ViralBShah
Copy link
Contributor Author

I am traveling, and don't have access to that computer where I need to
try it out. I should be able to test it out next week and get back to
you.

-viral

On Mon, Feb 13, 2012 at 4:53 PM, Xianyi Zhang
reply@reply.github.com
wrote:

Hi Viral,

I just added those CPUs. You can download it on develop branch.
Could you help me test it?

Thanks
Xianyi


Reply to this email directly or view it on GitHub:
#69 (comment)

-viral

@xianyi
Copy link
Collaborator

xianyi commented Feb 13, 2012

OK. I have tested it on my core i5. It works fine.

Xianyi

Viral B. Shah write:Sent: 2012/2/13 22:21

I am traveling, and don't have access to that computer where I need to
try it out. I should be able to test it out next week and get back to
you.

-viral

On Mon, Feb 13, 2012 at 4:53 PM, Xianyi Zhang
reply@reply.github.com
wrote:

Hi Viral,

I just added those CPUs. You can download it on develop branch.
Could you help me test it?

Thanks
Xianyi


Reply to this email directly or view it on GitHub:
#69 (comment)

@ViralBShah
Copy link
Contributor Author

A request - is it possible for you to make a release with this? This is the number one reason the julia build fails right now.

@xianyi
Copy link
Collaborator

xianyi commented Feb 19, 2012

Hi Viral,

Did this patch solve the problem?

If it solves the issue, I will release a version today.

Xianyi

Viral B. Shah write:Sent: 12-2-19 Morning 11:47

A request - is it possible for you to make a release with this? This is the number one reason the julia build fails right now.


Reply to this email directly or view it on GitHub:
#69 (comment)

@ViralBShah
Copy link
Contributor Author

I finally could try it out. It works great. Would be great if you can release a new version. Thanks!

@xianyi
Copy link
Collaborator

xianyi commented Feb 20, 2012

Hi Viral,

I have released the new version, 0.1 alpha 2.5

Xianyi

Viral B. Shah write:Sent: 12-2-20 Morning 11:47

I finally could try it out. It works great. Would be great if you can release a new version. Thanks!


Reply to this email directly or view it on GitHub:
#69 (comment)

@ViralBShah
Copy link
Contributor Author

Thanks much!

-viral

On 20-Feb-2012, at 10:02 AM, Xianyi Zhang wrote:

Hi Viral,

I have released the new version, 0.1 alpha 2.5

Xianyi

Viral B. Shah write:Sent: 12-2-20 Morning 11:47

I finally could try it out. It works great. Would be great if you can release a new version. Thanks!


Reply to this email directly or view it on GitHub:
#69 (comment)


Reply to this email directly or view it on GitHub:
#69 (comment)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants