GitLab

In particular, this allows us to split the “test” and “pkg” stages, without
increasing the execution time.

Test for ifndef

See merge request Dynare/dynare!1763

See Dynare/dynare#1747

— Fix the interpretation of @#else clause within an @#ifndef
  Closes: #1747
— Fix the interpretation of @#elseif clauses within an @#ifdef or @#ifndef
— PAC model: fixes to M_.pac.MODEL.ar (information about autoregressive part)
  Ref. preprocessor#52
— PAC model: make detection of non-optimizing part more robust
  Ref. preprocessor#50
— Fixes to Julia output

This bug is causing random crashes in the testsuite on our runners.

See https://sourceware.org/bugzilla/show_bug.cgi?id=19329

Bug introduced in commit e21cb1ad

Ref. #1720

This was causing some tests to be skipped.

The bug had been introduced e7162a2a.

Closes: #1720

The fix in commit 24cc67e5 was incorrect.

The routines use the find() function applied to a subset of columns of the
Jacobian, which in this case is a row vector. When passed a row vector, find()
returns row vectors (while it returns column vectors when passed a column
vector or a matrix). This case was not correctly handled.

Closes: #1735

It’s actually useless and complicates the code. Simply rely on
options_.bytecode.

Ref. #1742

Ref. #1742

By the way, document and enforce the fact that stochastic extended_path (i.e.
order > 0) is not compatible with either bytecode or block.

Closes: #1742

Ref. #1726

Ref. #1726

It is thus no longer needed to pass the MATLAB_VERSION option to the configure
script when compiling from source.

— simplify external function C output
— internal change to SymbolTable::getTypeSpecificID()
— ensure that negative constants are never marked as temporary terms

Closes: #1727

After simulating a block containing purely forward variables (thus of type
“evaluate backward”), the it_ variable of the evaluator would be left in an
inconsistent state (typically 0, which means that taking the value of a lagged
variable would lead to an invalid read).

By the way, fix a symmetric problem for backward blocks (which could
potentially create a invalid read for purely backward models).

Ref. #1727