-
Notifications
You must be signed in to change notification settings - Fork 34
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
Feature request: enable automated building #13
Comments
I would have no objection the suggested change. However, I was also
wondering if an alternative solution is to have one specific Makefile
that is customized for QUDA and that could be the one copied to the
directory in which MILC is being compiled.
…On Fri, 2017-07-14 at 16:57 +0000, maddyscientist wrote:
At present the MILC build system requires manual copying and editing
of the Makefile to set building options, e.g., whether to build with
QUDA of QPhiX support. This prevents automated testing of MILC.
While using cmake or even autotools would be the most user friendly,
and most desired from our perspective, I note that simply extending
the current Makefile to support environment variables for setting
paths would give the rudimentary support we would like.
E.g., from the current Makefile, there is:
WANTQUDA = true
WANT_CL_BCG_GPU = false
WANT_FN_CG_GPU = true
WANT_FL_GPU = true
WANT_FF_GPU = true
WANT_GF_GPU = true
If this was simply changed to
WANTQUDA ?= true
WANT_CL_BCG_GPU ?= false
WANT_FN_CG_GPU ?= true
WANT_FL_GPU ?= true
WANT_FF_GPU ?= true
WANT_GF_GPU ?= true
then the user could set these variables with environment variables,
overriding the contents of the Makefile. This could easily be extended
to the QIO, QMP, and other options as needed.
Background: we ***@***.*** and I) would like to enable QUDA's pull
request system to fetch the latest develop version of MILC, build
QUDA-enabled MILC and run MILC's internal rhmc and rhmd tests for each
pull request for QUDA. This will prevent any regressions in MILC
compatibility as QUDA continues to rapidly evolve and will prevent
wasted cycles going forward.
Thoughts?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.
|
Hi Kate and Steve,
@kate: Your suggestion is fine. I'll add that to my list for future
improvements. I know the MILC build system is clunky. @Steve: I would
prefer not to have a proliferation of official custom Makefiles -- it
would multiply the maintenance effort, since there would be a lot of
duplication between them that would then require duplicated updates.
Best,
Carleton
…On 07/14/2017 11:10 AM, stevengottlieb wrote:
I would have no objection the suggested change. However, I was also
wondering if an alternative solution is to have one specific Makefile
that is customized for QUDA and that could be the one copied to the
directory in which MILC is being compiled.
On Fri, 2017-07-14 at 16:57 +0000, maddyscientist wrote:
> At present the MILC build system requires manual copying and editing
> of the Makefile to set building options, e.g., whether to build with
> QUDA of QPhiX support. This prevents automated testing of MILC.
>
> While using cmake or even autotools would be the most user friendly,
> and most desired from our perspective, I note that simply extending
> the current Makefile to support environment variables for setting
> paths would give the rudimentary support we would like.
>
> E.g., from the current Makefile, there is:
>
> WANTQUDA = true
> WANT_CL_BCG_GPU = false
> WANT_FN_CG_GPU = true
> WANT_FL_GPU = true
> WANT_FF_GPU = true
> WANT_GF_GPU = true
> If this was simply changed to
>
> WANTQUDA ?= true
> WANT_CL_BCG_GPU ?= false
> WANT_FN_CG_GPU ?= true
> WANT_FL_GPU ?= true
> WANT_FF_GPU ?= true
> WANT_GF_GPU ?= true
> then the user could set these variables with environment variables,
> overriding the contents of the Makefile. This could easily be extended
> to the QIO, QMP, and other options as needed.
>
> Background: we ***@***.*** and I) would like to enable QUDA's pull
> request system to fetch the latest develop version of MILC, build
> QUDA-enabled MILC and run MILC's internal rhmc and rhmd tests for each
> pull request for QUDA. This will prevent any regressions in MILC
> compatibility as QUDA continues to rapidly evolve and will prevent
> wasted cycles going forward.
>
> Thoughts?
>
> —
> You are receiving this because you are subscribed to this thread.
> Reply to this email directly, view it on GitHub, or mute the thread.
>
>
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#13 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AF_j3mD1eT5PXQvMEXMRqdMQjdMrqzl5ks5sN6EVgaJpZM4OYfUE>.
|
Thanks Carleton. That will make scripting the MILC build a lot easier. |
Done (for QUDA macros) and pushed to milc_qcd/develop.
…On 7/14/17 10:57 AM, maddyscientist wrote:
At present the MILC build system requires manual copying and editing
of the Makefile to set building options, e.g., whether to build with
QUDA of QPhiX support. This prevents automated testing of MILC.
While using cmake or even autotools would be the most user friendly,
and most desired from our perspective, I note that simply extending
the current Makefile to support environment variables for setting
paths would give the rudimentary support we would like.
E.g., from the current Makefile, there is:
WANTQUDA =true
WANT_CL_BCG_GPU =false
WANT_FN_CG_GPU =true
WANT_FL_GPU =true
WANT_FF_GPU =true
WANT_GF_GPU =true
If this was simply changed to
WANTQUDA?=true
WANT_CL_BCG_GPU?=false
WANT_FN_CG_GPU?=true
WANT_FL_GPU?=true
WANT_FF_GPU?=true
WANT_GF_GPU?=true
then the user could set these variables with environment variables,
overriding the contents of the Makefile. This could easily be extended
to the QIO, QMP, and other options as needed.
Background: we ***@***.*** <https://github.com/mathiaswagner> and
I) would like to enable QUDA's pull request system to fetch the latest
develop version of MILC, build QUDA-enabled MILC and run MILC's
internal rhmc and rhmd tests for each pull request for QUDA. This will
prevent any regressions in MILC compatibility as QUDA continues to
rapidly evolve and will prevent wasted cycles going forward.
Thoughts?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#13>, or mute the thread
<https://github.com/notifications/unsubscribe-auth/AF_j3jgWBBBuvFybezNY3W1rufYG857Dks5sN54JgaJpZM4OYfUE>.
|
detar
pushed a commit
that referenced
this issue
Mar 29, 2019
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
At present the MILC build system requires manual copying and editing of the Makefile to set building options, e.g., whether to build with QUDA of QPhiX support. This prevents automated testing of MILC.
While using cmake or even autotools would be the most user friendly, and most desired from our perspective, I note that simply extending the current Makefile to support environment variables for setting paths would give the rudimentary support we would like.
E.g., from the current Makefile, there is:
If this was simply changed to
then the user could set these variables with environment variables, overriding the contents of the Makefile. This could easily be extended to the QIO, QMP, and other options as needed.
Background: we (@mathiaswagner and I) would like to enable QUDA's pull request system to fetch the latest develop version of MILC, build QUDA-enabled MILC and run MILC's internal rhmc and rhmd tests for each pull request for QUDA. This will prevent any regressions in MILC compatibility as QUDA continues to rapidly evolve and will prevent wasted cycles going forward.
Thoughts?
The text was updated successfully, but these errors were encountered: