Sven Verdoolaege [Wed, 13 Oct 2010 17:39:35 +0000 (19:39 +0200)]
isl_mat_right_inverse: fix memory leak on error path
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 11 Oct 2010 11:14:12 +0000 (13:14 +0200)]
isl_union_pw_qpolynomial_zero: check for NULL input
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 2 Oct 2010 13:33:55 +0000 (15:33 +0200)]
export isl_basic_set_sample_point
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 2 Oct 2010 10:34:01 +0000 (12:34 +0200)]
basis_reduction_templ.c: fix typo in comment
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 2 Oct 2010 09:16:17 +0000 (11:16 +0200)]
export isl_basic_set_reduced_basis
Since this function will now be available from the outside, we
add some more checks and we allow the input to contain equalities.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 2 Oct 2010 09:12:52 +0000 (11:12 +0200)]
add isl_int_[gs]et_gmp
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 2 Oct 2010 08:51:14 +0000 (10:51 +0200)]
isl_pw_qpolynomial_print: print modulo constraints in ISL_FORMAT_C
In particular, print "i % 2 == 0" instead of "i - 2 * floord(i,2) == 0".
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 1 Oct 2010 15:05:16 +0000 (17:05 +0200)]
export isl_basic_set_lift
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 30 Sep 2010 14:15:36 +0000 (16:15 +0200)]
add isl_union_map_intersect_range
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 30 Sep 2010 08:26:09 +0000 (10:26 +0200)]
add isl_constraint_get_dim_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 30 Sep 2010 07:52:19 +0000 (09:52 +0200)]
add isl_set_n_basic_set
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 17:03:55 +0000 (19:03 +0200)]
add isl_pw_qpolynomial_set_dim_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 15:57:36 +0000 (17:57 +0200)]
add isl_union_pw_qpolynomial_extract_pw_qpolynomial
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 15:38:43 +0000 (17:38 +0200)]
add isl_union_set_extract_set
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 15:25:39 +0000 (17:25 +0200)]
add isl_union_set_n_set
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 10:52:16 +0000 (12:52 +0200)]
add isl_set_get_dim_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 29 Sep 2010 10:56:26 +0000 (12:56 +0200)]
add isl_set_get_tuple_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 27 Sep 2010 08:24:35 +0000 (10:24 +0200)]
add isl_union_set_dump for debugging
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 30 Sep 2010 13:47:43 +0000 (15:47 +0200)]
rename isl_map_add to isl_map_add_dims
The new name is more consistent with other functions and avoids confusion
as to what is being added.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 27 Sep 2010 08:47:25 +0000 (10:47 +0200)]
isl_basic_map_gauss: normalize resulting constraints
When isl_basic_map_gauss is called from isl_basic_map_simplify,
the constraints get normalized already. However, sometimes the calling
functions wants to only perform isl_basic_map_gauss and then the constraints
should still be normalized.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 30 Sep 2010 09:46:09 +0000 (11:46 +0200)]
fix return type of isl_union_set_lex_*_union_set
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 4 Oct 2010 20:08:54 +0000 (22:08 +0200)]
isl_basic_set_multiplicative_call: handle parametric input
Due to a mistake in the original implementation, the wrong dimensions
would get dropped from factors of the input in case this input was
parametric.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sun, 3 Oct 2010 22:35:23 +0000 (00:35 +0200)]
isl_dim_replace: call isl_dim_cow before changing isl_dim
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 25 Sep 2010 13:51:53 +0000 (15:51 +0200)]
export isl_access_info_free
It was thought before that there was no need to export this function,
because it is called implicitly by isl_access_info_compute_flow,
but if some error occurs in the application before the call
to isl_access_info_compute_flow, isl_access_info_free should be
called to free the (partially constructed) isl_access_info.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 20:29:10 +0000 (22:29 +0200)]
add isl_basic_map_get_tuple_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 14:51:08 +0000 (16:51 +0200)]
add isl_basic_map_remove_divs
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 08:10:28 +0000 (10:10 +0200)]
add isl_union_map_compute_flow
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 11:06:55 +0000 (13:06 +0200)]
isl_access_info_compute_flow: preserve dimension specifications
The manipulations on the sink map would destroy the internal
structure of the domain of that map. By using the appropriate
functions, we can avoid this loss of information.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 08:09:52 +0000 (10:09 +0200)]
doc: document type of isl_access_info_alloc callback function
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 06:48:24 +0000 (08:48 +0200)]
add isl_union_map_range_map and isl_union_map_domain_map
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 06:45:19 +0000 (08:45 +0200)]
add isl_basic_map_from_domain
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 24 Sep 2010 05:38:50 +0000 (07:38 +0200)]
add isl_union_map_apply_domain
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 23 Sep 2010 19:49:24 +0000 (21:49 +0200)]
add isl_dim_from_domain and isl_dim_from_range
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 25 Sep 2010 12:51:24 +0000 (14:51 +0200)]
doc: document some isl_dim manipulation functions
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 23 Sep 2010 19:16:48 +0000 (21:16 +0200)]
add isl_set_set_tuple_name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 21 Jul 2010 19:19:11 +0000 (21:19 +0200)]
add isl_basic_{set,map}_from_constraint_matrices
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 9 Feb 2010 09:10:48 +0000 (10:10 +0100)]
add isl_basic_map_{in,}equalities_matrix
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 21 Sep 2010 09:20:45 +0000 (11:20 +0200)]
isl_input.c: accept divs in affine expressions
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 22 Sep 2010 19:58:19 +0000 (21:58 +0200)]
Revert "isl_input.c: read_tuple: construct isl_dim instead of isl_basic_map"
This reverts commit
6373465b50e93c0a8df20fdefe572a2eaebb8490.
It was thought that handling nested dims would be easier when
constructin isl_dims instead of isl_basic_maps during construction.
However, it turns out that using isl_basic_maps works just fine
and avoids carrying around the equalities separately.
Moreover, if we want to be able to parse integer divisions,
then we need to store their definitions somewhere and
an isl_basic_map is the most appropriate way to do so.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 22 Sep 2010 08:01:54 +0000 (10:01 +0200)]
isl_input.c: obj_read: error out on intermediate NULL results
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 20 Sep 2010 17:03:10 +0000 (19:03 +0200)]
add isl_vec_zero_extend
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 20 Sep 2010 16:41:44 +0000 (18:41 +0200)]
isl_input.c: store div definition in variable
We want to be able to handle divs inside affine expressions and
so we need to store the definition of an existentially quantified
variable somewhere.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 22 Sep 2010 09:05:26 +0000 (11:05 +0200)]
isl_basic_set_swap_vars: finalize result
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 21 Sep 2010 11:53:16 +0000 (13:53 +0200)]
isl_printer_print_qpolynomial: put parentheses around lone integer division
Otherwise, the integer division could be mistaken for a tuple.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 14 Sep 2010 03:53:30 +0000 (05:53 +0200)]
add isl_basic_set_multiplicative_call
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 16 Sep 2010 01:53:30 +0000 (03:53 +0200)]
add isl_qpolynomial_one
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 14 Sep 2010 04:29:03 +0000 (06:29 +0200)]
merge isl_basic_set_drop_constraints_involving implementations
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 14 Sep 2010 03:53:27 +0000 (05:53 +0200)]
add isl_basic_map_from_range
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 11 Sep 2010 17:59:59 +0000 (19:59 +0200)]
fix parameter alignment when alignee has zero parameters
The number of parameters was passed as an unsigned integer and
later decremented, leading to an allocation request for a huge
amount of memory.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 10 Sep 2010 11:06:15 +0000 (13:06 +0200)]
isl 0.04
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 8 Sep 2010 18:18:10 +0000 (20:18 +0200)]
add isl_union_set_lex_[lg][et]_union_set
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 9 Sep 2010 19:28:45 +0000 (21:28 +0200)]
isl_dim_replace: recursively replace parameters in nested dims
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 4 Sep 2010 12:45:01 +0000 (14:45 +0200)]
isl_input.c add_equalities: gauss and finalize result after adding equalities
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sun, 15 Aug 2010 13:06:26 +0000 (15:06 +0200)]
add public API for matrices
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 2 Sep 2010 15:15:41 +0000 (17:15 +0200)]
isl_coalesce.c: fix typos in comment
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 1 Sep 2010 14:35:56 +0000 (16:35 +0200)]
isl_polytope_scan: improve error handling
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 31 Aug 2010 09:23:20 +0000 (11:23 +0200)]
add isl_pw_qpolynomial_split_periods
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sun, 29 Aug 2010 21:05:29 +0000 (23:05 +0200)]
isl_polynomial.c: sort_divs: remove duplicate divs, if any
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sun, 29 Aug 2010 15:34:35 +0000 (17:34 +0200)]
isl_polynomial.c: sort_divs: fix ordering of div rows
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 26 Aug 2010 15:29:35 +0000 (17:29 +0200)]
argument parsing: add ISL_ARG_USER_OPT_CHOICE
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 24 Aug 2010 19:01:20 +0000 (21:01 +0200)]
align parameters of union arguments to binary functions
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 24 Aug 2010 18:59:04 +0000 (20:59 +0200)]
add isl_dim_replace
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 24 Aug 2010 18:53:12 +0000 (20:53 +0200)]
add isl_union_*_dump for debugging
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 26 Aug 2010 12:24:13 +0000 (14:24 +0200)]
clean up isl_pw_qpolynomial_neg
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 26 Aug 2010 08:19:30 +0000 (10:19 +0200)]
isl_union_templ.c: minor clean-up
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 25 Aug 2010 15:59:39 +0000 (17:59 +0200)]
add isl_mat_insert_zero_cols
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 25 Aug 2010 18:56:54 +0000 (20:56 +0200)]
add isl_basic_map_get_ctx
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 24 Aug 2010 08:02:34 +0000 (10:02 +0200)]
isl_map_read_from_str: parse tuple entries with leading minus sign
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 18 Aug 2010 11:12:31 +0000 (13:12 +0200)]
add isl_map_apply_pw_qpolynomial_fold
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 18 Aug 2010 11:09:29 +0000 (13:09 +0200)]
add isl_union_pw_qpolynomial_fold_get_type
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 18 Aug 2010 10:24:28 +0000 (12:24 +0200)]
add isl_pw_qpolynomial_fold_insert_dims
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 17 Aug 2010 19:04:48 +0000 (21:04 +0200)]
add isl_union_pw_qpolynomial_fold_add_union_pw_qpolynomial
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 17 Aug 2010 11:43:11 +0000 (13:43 +0200)]
add type to isl_{union_,}pw_qpolynomial_fold
The inner isl_qpolynomial_folds should all be of the same type,
so we need to keep track of the type at the outer level.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 17 Aug 2010 07:47:21 +0000 (09:47 +0200)]
fix isl_qpolynomial_fold_reset_dim
In particular, the old version wouldn't cow and it wouldn't change
the dimension of the nested isl_qpolynomials.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 16 Aug 2010 12:36:58 +0000 (14:36 +0200)]
isl_input.c: read_term: allow terms starting with negative constant
In particular, we used to fail to parse "1-3x" because -3 is treated
as a negative constant. Now, we treat is in the same way as "1+-3x".
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 7 Aug 2010 13:26:11 +0000 (15:26 +0200)]
isl_union_pw_qpolynomial_add_pw_qpolynomial: don't add empty parts
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 7 Aug 2010 11:49:46 +0000 (13:49 +0200)]
isl_stream_read_obj: read reductions
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 7 Aug 2010 11:47:57 +0000 (13:47 +0200)]
isl_input.c: read_factor: fix read after free in error path
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 7 Aug 2010 11:45:38 +0000 (13:45 +0200)]
isl_stream_error: print unexpected identifier name
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 18 Aug 2010 08:11:25 +0000 (10:11 +0200)]
rename isl_pw_qpolynomial_fold_add to isl_pw_qpolynomial_fold_fold
The new name is more consistent with the naming of
isl_qpolynomial_fold_fold. Furthermore, it makes sense to perform
addition on isl_pw_qpolynomial_folds, so we prefer to use
the name isl_pw_qpolynomial_fold_add for such a function.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sun, 15 Aug 2010 11:16:41 +0000 (13:16 +0200)]
doc: minor clarification on operations on union sets
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Sat, 14 Aug 2010 16:48:26 +0000 (18:48 +0200)]
add isl_union_set_get_ctx
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 11 Aug 2010 10:49:05 +0000 (12:49 +0200)]
isl_map_gist: also perform gist in universe context
Originally, it didn't make sense to gist in a universe context,
because there were no constraints with respect to which the gist
could be computed. However, since the more aggressive gist of
d5e1701 (isl_basic_map_gist: be more aggressive in removing constraints)
some contraints can get removed even in a universe context.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 10 Aug 2010 18:50:22 +0000 (20:50 +0200)]
isl_union_map: construct new isl_union_map on operations that change dimensions
Otherwise, the entries in the hashtable will end up hashed on the wrong
hash value (that of the original isl_dim).
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 10 Aug 2010 18:03:28 +0000 (20:03 +0200)]
isl_union_map_is_empty: fix typo that resulted in memory leaks
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 9 Aug 2010 15:07:10 +0000 (17:07 +0200)]
add isl_basic_map_has_defining_equality
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 9 Aug 2010 12:08:30 +0000 (14:08 +0200)]
add isl_set_get_ctx
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Uday Kumar Reddy [Sun, 8 Aug 2010 15:14:39 +0000 (11:14 -0400)]
update .gitignore
Sven Verdoolaege [Sat, 7 Aug 2010 15:43:03 +0000 (17:43 +0200)]
isl_printer_print_map: fix printing in latex format
It was broken by e094707 (isl_output.c: extract out common print_dim).
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 6 Aug 2010 13:25:01 +0000 (15:25 +0200)]
isl_set_wrap_facet: normalize resulting facet constraint
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 6 Aug 2010 13:24:32 +0000 (15:24 +0200)]
isl_coalesce.c: fix typo in comment
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 6 Aug 2010 09:34:26 +0000 (11:34 +0200)]
isl_pw_qpolynomial_intersect_domain: simplify polynomials using equalities
We now use any equalities we find in the intersected domain.
Perhaps we should check if the intersection had any effect
on the equalities first.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 6 Aug 2010 10:28:04 +0000 (12:28 +0200)]
isl_qpolynomial_move_dims: plug memory leak and fix reordering of divs
In particular, when divs got reordered, the new order would not
be reflected in the associated upoly.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Fri, 6 Aug 2010 09:46:33 +0000 (11:46 +0200)]
add isl_pw_qpolynomial_read_from_str
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 5 Aug 2010 18:10:12 +0000 (20:10 +0200)]
add isl_qpolynomial_fold_substitute
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 5 Aug 2010 18:01:08 +0000 (20:01 +0200)]
isl_pw_qpolynomial_intersect_domain: drop pieces that end up having empty domain
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Tue, 3 Aug 2010 12:46:48 +0000 (14:46 +0200)]
isl_map_product: used nested spaces in result
That is, the product of A -> B and C -> D now has dimension
[A -> C] -> [B -> D]. This allows us to keep all information
about the spaces A, B, C, D, including their names and any
prior nestings.
Since it sometimes also useful to have concatenated spaces in the result,
we also introduct isl_map_flat_product, which produces what isl_map_product
used to produce.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Wed, 4 Aug 2010 19:52:34 +0000 (21:52 +0200)]
isl_pw_qpolynomial_bound: handle isl_pw_qpolynomials with wrapped domains
In particular, compute a bound over the range of the wrapped relation,
instead of over the entire relation.
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Mon, 2 Aug 2010 14:56:35 +0000 (16:56 +0200)]
support nested isl_dims
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>
Sven Verdoolaege [Thu, 5 Aug 2010 14:10:41 +0000 (16:10 +0200)]
doc: fix typo
Signed-off-by: Sven Verdoolaege <skimo@kotnet.org>