Searched hist:"0 e9bae810fdaeb60e2713eaa8ddb89f42e079fd1" (Results 1 – 4 of 4) sorted by relevance
| /petsc/src/sys/error/ |
| H A D | adebug.c | 0e9bae810fdaeb60e2713eaa8ddb89f42e079fd1 Mon Sep 10 04:12:29 UTC 2012 Barry Smith <bsmith@mcs.anl.gov> MatSetValues(Blocked)_MPI(B)AIJ did not handle no new zero entry error generation in off diagonal part
MatSetValues(Blocked)Stencil stupidly used PETSC_MIN_INT to indicate a negative row/column (meaning not to insert anything there), but then MatSetValuesBlocked() multiplied these numbers by the block size the turning the negative number into a positive which MatSetValues() then tried to put into the matrix (even though it was a garbage location)
added support for using lldb debugger with -start_in_debugger and -on_error_attach_debugger
Hg-commit: 817ef0048927d966788e5fa1c4cdadf402d2ac09
|
| /petsc/src/mat/impls/baij/mpi/ |
| H A D | mpibaij.c | 0e9bae810fdaeb60e2713eaa8ddb89f42e079fd1 Mon Sep 10 04:12:29 UTC 2012 Barry Smith <bsmith@mcs.anl.gov> MatSetValues(Blocked)_MPI(B)AIJ did not handle no new zero entry error generation in off diagonal part
MatSetValues(Blocked)Stencil stupidly used PETSC_MIN_INT to indicate a negative row/column (meaning not to insert anything there), but then MatSetValuesBlocked() multiplied these numbers by the block size the turning the negative number into a positive which MatSetValues() then tried to put into the matrix (even though it was a garbage location)
added support for using lldb debugger with -start_in_debugger and -on_error_attach_debugger
Hg-commit: 817ef0048927d966788e5fa1c4cdadf402d2ac09
|
| /petsc/src/mat/impls/aij/mpi/ |
| H A D | mpiaij.c | 0e9bae810fdaeb60e2713eaa8ddb89f42e079fd1 Mon Sep 10 04:12:29 UTC 2012 Barry Smith <bsmith@mcs.anl.gov> MatSetValues(Blocked)_MPI(B)AIJ did not handle no new zero entry error generation in off diagonal part
MatSetValues(Blocked)Stencil stupidly used PETSC_MIN_INT to indicate a negative row/column (meaning not to insert anything there), but then MatSetValuesBlocked() multiplied these numbers by the block size the turning the negative number into a positive which MatSetValues() then tried to put into the matrix (even though it was a garbage location)
added support for using lldb debugger with -start_in_debugger and -on_error_attach_debugger
Hg-commit: 817ef0048927d966788e5fa1c4cdadf402d2ac09
|
| /petsc/src/mat/interface/ |
| H A D | matrix.c | 0e9bae810fdaeb60e2713eaa8ddb89f42e079fd1 Mon Sep 10 04:12:29 UTC 2012 Barry Smith <bsmith@mcs.anl.gov> MatSetValues(Blocked)_MPI(B)AIJ did not handle no new zero entry error generation in off diagonal part
MatSetValues(Blocked)Stencil stupidly used PETSC_MIN_INT to indicate a negative row/column (meaning not to insert anything there), but then MatSetValuesBlocked() multiplied these numbers by the block size the turning the negative number into a positive which MatSetValues() then tried to put into the matrix (even though it was a garbage location)
added support for using lldb debugger with -start_in_debugger and -on_error_attach_debugger
Hg-commit: 817ef0048927d966788e5fa1c4cdadf402d2ac09
|