ADMB for raspberry pi 4

22 views
Skip to first unread message

Ernesto Jardim

unread,
May 8, 2020, 1:58:26 PM5/8/20
to us...@admb-project.org
Hi,

I'm trying to compile ADMB in a raspberry pi 4 with Ubuntu 19.10 and gcc 9.2.1 . The architecture is arm 64. I'm getting this error both with 11.6 and 12.1. Any ideas how to fix or workaround.

Thanks

Best

EJ

df1b2-separable/df1b2f24.cpp: In member function ‘virtual void df1b2_init_bounded_matrix::set_value(const init_df1b2vector&, const int&, const df1b2variable&)’:
df1b2-separable/df1b2f24.cpp:145:41: error: cast from ‘df1b2variable*’ to ‘int’ loses precision [-fpermissive]
  145 |         int tmp= (int)( &((*this)(i,j)) );

Johnoel

unread,
May 8, 2020, 3:00:03 PM5/8/20
to Ernesto Jardim, ADMB Users
This would be really neat to port to Raspberry Pi.  ADMB has not been built for the ARM 64 processor.
An issue has been set up to fix it.

A possible workaround is to force a 64-bit build using the command below

"in admb root directory" $ make CXXFLAGS=-D__x86_64

--
To unsubscribe from this group and stop receiving emails from it, send an email to users+un...@admb-project.org.

chrisg...@gmail.com

unread,
May 11, 2020, 5:22:45 PM5/11/20
to ADMB Users, ernesto...@gmail.com
As it mentions, you could just add -fpermissive to the compiler flags to make these errors become warnings.
Chris


On Friday, May 8, 2020 at 12:00:03 PM UTC-7, johnoel wrote:
This would be really neat to port to Raspberry Pi.  ADMB has not been built for the ARM 64 processor.
An issue has been set up to fix it.

A possible workaround is to force a 64-bit build using the command below

"in admb root directory" $ make CXXFLAGS=-D__x86_64

On Fri, May 8, 2020 at 7:58 AM Ernesto Jardim <ernest...@gmail.com> wrote:
Hi,

I'm trying to compile ADMB in a raspberry pi 4 with Ubuntu 19.10 and gcc 9.2.1 . The architecture is arm 64. I'm getting this error both with 11.6 and 12.1. Any ideas how to fix or workaround.

Thanks

Best

EJ

df1b2-separable/df1b2f24.cpp: In member function ‘virtual void df1b2_init_bounded_matrix::set_value(const init_df1b2vector&, const int&, const df1b2variable&)’:
df1b2-separable/df1b2f24.cpp:145:41: error: cast from ‘df1b2variable*’ to ‘int’ loses precision [-fpermissive]
  145 |         int tmp= (int)( &((*this)(i,j)) );

--
To unsubscribe from this group and stop receiving emails from it, send an email to us...@admb-project.org.

johnoel

unread,
Mar 12, 2021, 11:03:53 PM3/12/21
to ADMB Users, Ernesto Jardim, us...@admb-project.org
Hi Ernesto, 

I just noticed you tried to build for arm64 with linux.

The latest version of ADMB-12.3 should work with arm64 chips.  Use the
procedure in the link below to download and build ADMB-12.3.


Please let me know if it works.


Reply all
Reply to author
Forward
0 new messages