Jack Howarth wrote:

Roberto, I'll try that tonight and post the results.

OK.

I am a bit confused that ppl hasn't been tested on Darwin since I am pretty sure Sebastian Pop said that graphite had been tested on that platform (unless that was only with polylib instead).

I didn't say it hasn't been tested: we have several users who use PPL on Darwin and any bug that has been reported has been fixed. The point is that, right now, no PPL developer has access to such a machine: we thus depend on the input our users provide to us. All the best,

Roberto

Jack Howarth wrote:

 Do you mean that everything has been tested on PowerPC

Darwin or that everything has been tested under Tiger (darwin8) rather than Leopard (darwin9)? I'll try building ppl 0.9 under powerpc-apple-darwin9 here and see what the Make check results are for that target. Jack ps I can also try using the gcc-4.2 compilers in Xcode 3.1 to see if that makes a difference on i686-apple-darwin9. Oh, did the folks who tested ppl on Darwin use Apple's gcc compilers or the FSF ones?

Dear Jack,

I don't have the information you would like to have. The only thing I can say is that we will work with you in order to fix any bug you may identify. And if someone provides us with access to some darwinX machines we can use for regression-testing... Cheers,

Roberto

Jack Howarth wrote:

I just noticed that the gmp package in fink doesn't

build gmp with the -fexception configure option. Do you think that could cause make check failures in ppl 0.9?

No: the configure script should detect this and disable the programs that test recovery from out-of-memory conditions.

Are there any instructions for the proper sequence of commands to generate the configure required to build ppl cvs? I was going to try that to make sure that the Macintel testsuite failures weren't already fixed in ppl cvs but it is unclear how to make the ppl cvs checkout directory buildable. Thanks in advance. Jack

Roberto, I have verified that there are no make check failures in ppl-0.9 on powerpc-apple-darwin9 with Xcode 3.1. The problems appear to be specific to Macintel. I see the same 5 failures when ppl-0.9 is built with Apple's gcc 4.0.1 or gcc 4.2 compilers (so it is unlikely to be a compiler bug). The failing test cases output is appended below. Jack

=== test01 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(x, y) *** B <= 2, B - A == 0

=== test02 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(x, x + 4) *** A <= 6, B <= 2, A - B <= 7

=== test03 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(x, 4) *** A == 4, B <= 2

=== test04 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(x, x) *** A <= 2, B <= 2, A - B <= 3

=== test05 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(x, 2*x - 2, 2) *** A <= 1, B <= 2, A - B <= 2

=== test06 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(y, 2*x, 2) *** A <= 2, B - A == 0

=== test07 === *** bd1 *** A <= 2, B <= 2, A - B <= 3 *** bd1.affine_image(y, 3*x + 3, 3) *** A <= 2, B - A == 1

=== test08 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 Containment does not hold: computed result is A >= -13421773/16777216, A <= 11744051/8388608, B >= -1, B <= 2, A - B <= 5033165/8388608, B - A <= 1 known result is A >= -4/5, A <= 7/5, B >= -1, B <= 2, A - B <= 3/5, B - A <= 1 *** bd.affine_image(x, -2*x - 3*y + 1, -5) *** A >= -0.800000011920928955078125, A <= 1.39999997615814208984375, B >= -1, B <= 2, A - B <= 0.60000002384185791015625, B - A <= 1

=== test09 === *** bd *** A <= 1, B <= 2, C >= 3 *** bd.affine_image(z, x + 2*y -3*z + 2, 4) *** A <= 1, B <= 2, C <= -0.5

=== test10 === *** bd *** A <= 1, B >= 1, B <= 2, C <= 0, D == 3 *** bd.affine_image(A, -B + 2*C + 1, -3) *** A >= 0, B >= 1, B <= 2, C <= 0, D == 3, B - A <= 1.6666667461395263671875, B - C >= 1, D - B >= 1, D - B <= 2, D - C >= 3

=== test11 === *** bd *** A <= 1, B <= 2 *** bd.affine_image(x, 2*y + z + 2, 4) *** B <= 2

=== test12 === invalid_argument: PPL::BD_Shape::affine_image(v, e, d): d == 0

=== test13 === invalid_argument: PPL::BD_Shape::affine_image(v, e, d): this->space_dimension() == 1, required dimension == 1.

=== test14 === invalid_argument: PPL::BD_Shape::affine_image(v, e, d): this->space_dimension() == 2, e->space_dimension() == 3.

=== test15 ===

=== test16 === failed tests: test08 FAIL: affineimage1

=== test01 === *** bd *** A <= 1, B <= 2, C >= 3 *** bd.affine_image(x, -x) *** A >= -1, B <= 2, C >= 3, C - B >= 1

=== test02 === *** bd *** A <= 1, B <= 2, C >= 3 *** bd.affine_image(x, -z) *** A <= -3, B <= 2, C >= 3, C - B >= 1

=== test03 === *** bd *** A <= 1, B >= 1, B <= 2 *** bd.affine_image(x, -y + 1) *** A >= -1, A <= 0, B >= 1, B <= 2

=== test04 === *** bd *** A <= 1, B >= 1, B <= 2 *** bd.affine_image(x, -2*y + 1, -2) *** B >= 1, B <= 2, B - A == 0.5

=== test05 === *** bd *** A <= 1, B >= 1, B <= 2 *** bd.affine_image(x, -2*y + 1, 2) *** A >= -1.5, A <= -0.5, B >= 1, B <= 2

=== test06 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 *** bd.affine_image(x, 2*x + y + 1) *** A >= 0, A <= 5, B >= -1, B <= 2, A - B <= 3, A - B >= 1

=== test07 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 *** bd.affine_image(x, -2*x + y + 1) *** A >= -2, A <= 3, B >= -1, B <= 2, A - B <= 1, B - A <= 1

=== test08 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 *** bd.affine_image(x, 2*x - 3*y + 1, 5) *** A >= -1, A <= 1.2000000476837158203125, B >= -1, B <= 2

=== test09 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 Containment does not hold: computed result is A >= -11744051/8388608, A <= 13421773/16777216, B >= -1, B <= 2 known result is A >= -7/5, A <= 4/5, B >= -1, B <= 2, A - B <= 9/5, B - A <= 17/5 *** bd.affine_image(x, -2*x - 3*y + 1, 5) *** A >= -1.39999997615814208984375, A <= 0.800000011920928955078125, B >= -1, B <= 2

=== test10 === *** bd *** A >= 0, A <= 1, B >= -1, B <= 2 *** bd.affine_image(x, 2*x - 3*y + 1, -5) *** A >= -1.2000000476837158203125, A <= 1, B >= -1, B <= 2, A - B <= 0.2000000476837158203125, B - A <= 1.400000095367431640625

=== test11 === *** bd *** B >= 0, B <= 2, C <= 3 Containment does not hold: computed result is A <= 11883861/2097152, B >= 0, B <= 2, C <= 3, A - B <= 5, A - C <= 5592405/2097152, C - B <= 3 known result is A <= 17/3, B >= 0, B <= 2, C <= 3, A - B <= 5, A - C <= 8/3, C - B <= 3 *** bd.affine_image(x, y + 5*z, 3) *** A <= 5.666666507720947265625, B >= 0, B <= 2, C <= 3, A - B <= 5, A - C <= 2.666666507720947265625, C - B <= 3 failed tests: test09 test11 FAIL: affineimage2

=== test01 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(A, 4*B + 6*C + 2, -2) *** B <= 0, C >= 0, B - C <= 0

=== test02 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(A, 2*A + 3*C + 2, 2) *** A <= -2, B <= 0, C >= 0, B - C <= 0

=== test03 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(A, -3*A + C - 1, 2) *** A <= -2, B <= 0, C >= 0, B - C <= 0

=== test04 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(A, 3*A + C - 1, -2) *** B <= 0, C >= 0, B - C <= 0

=== test05 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(A, -3*A + C - 1, -2) *** B <= 0, C >= 0, B - C <= 0

=== test06 === *** bd *** A == 2, B == 0, C >= -3, D <= 5 Containment does not hold: computed result is A == 2, B == 0, C >= -3, D <= 11744051/4194304, A - B == 2, A - C <= 5, B - C <= 3, D - B <= 11744051/4194304 known result is A == 2, B == 0, C >= -3, D <= 14/5, A - B == 2, A - C <= 5, D - A <= 4/5, B - C <= 3, D - B <= 14/5, D - C <= 29/5 *** bd.affine_preimage(D, 4*A - B + 2*C + 5*D - 1, 3) *** A == 2, B == 0, C >= -3, D <= 2.7999999523162841796875, A - B == 2, A - C <= 5, B - C <= 3, D - B <= 2.7999999523162841796875

=== test07 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(B, -B) *** A <= -1, B >= 0, C >= 0, C - A >= 1

=== test08 === *** bd *** A <= -1, B <= 0, C >= 0 *** bd.affine_preimage(B, -B, -1) *** A <= -1, B <= 0, C >= 0, C - A >= 1, B - C <= 0 failed tests: test06 FAIL: affinepreimage2

=== test01 === *** bd *** A >= -6, A <= 4, B == 0 *** bd.generalized_affine_image(y, LESS_THAN_OR_EQUAL, -y + 1) *** A >= -6, A <= 4, B <= 1

=== test02 === *** bd *** A >= -6, A <= 4, B == 0 *** bd.generalized_affine_image(x, GREATER_THAN_OR_EQUAL, -x - 3) *** A >= -7, B == 0

=== test03 === *** bd *** A >= 0, B <= 1 *** bd.generalized_affine_image(B, LESS_THAN_OR_EQUAL, 3*B + 1, 2) *** A >= 0, B <= 2

=== test04 === *** bd *** A == 0, B >= 1 *** bd.generalized_affine_image(B, GREATER_THAN_OR_EQUAL, B - 2) *** A == 0, B >= -1, A - B <= 1

=== test05 === *** bd *** B <= 1, B - A == 0 *** bd.generalized_affine_image(A, GREATER_THAN_OR_EQUAL, 2*A + 3, 2) *** B <= 1, A - B >= 1.5

=== test06 === *** bd *** B <= 1, B - A == 0, C - A <= 2 Containment does not hold: computed result is A <= 1, B <= 1, C <= 11744051/8388608, B - A == 0 known result is A <= 1, B <= 1, C <= 7/5, B - A == 0 *** bd.generalized_affine_image(C, LESS_THAN_OR_EQUAL, 2*C + 1, 5) *** A <= 1, B <= 1, C <= 1.39999997615814208984375, B - A == 0

=== test07 === *** bd *** B <= 1, B - A == 0, C - A <= 2 *** bd.generalized_affine_image(C, EQUAL, 5*C - 3, 4) *** A <= 1, B <= 1, C <= 3, B - A == 0

=== test08 === *** bd *** B <= 1, B - A == 0, C - A <= 2 *** bd.generalized_affine_image(B, GREATER_THAN_OR_EQUAL, -B - 2, 3) *** A <= 1, B >= -1, C <= 3, C - A <= 2

=== test09 === *** bd *** B <= 1, B - A == 0, C - A <= 2 *** bd.generalized_affine_image(B, LESS_THAN_OR_EQUAL, 4*A - 2*C + 3, -3) *** A <= 1, C <= 3, C - A <= 2

=== test10 === *** bd *** B <= 1, B - A == 0, C - A <= 2 *** bd.generalized_affine_image(B, EQUAL, 2*A - 4*B + C + 3, 3) *** A <= 1, C <= 3, C - A <= 2

=== test11 === *** bd *** A >= 0, A <= 4, B <= 5 *** bd.generalized_affine_image(A, EQUAL, 1) *** A == 1, B <= 5

=== test12 === *** bd *** A >= 0, A <= 4, B <= 5 *** bd.generalized_affine_image(B-3,EQUAL, B+1) *** A >= 0, A <= 4, B <= 9, B - A <= 9

=== test13 === invalid_argument: PPL::BD_Shape::generalized_affine_image(v, r, e, d): r is a strict relation symbol and *this is a BD_Shape

=== test14 === invalid_argument: PPL::BD_Shape::generalized_affine_image(v, r, e, d): r is a strict relation symbol and *this is a BD_Shape

=== test15 === invalid_argument: PPL::BD_Shape::generalized_affine_image(v, r, e, d): d == 0

=== test16 === invalid_argument: PPL::BD_Shape::generalized_affine_image(v, r, e, d): this->space_dimension() == 2, e->space_dimension() == 3.

=== test17 === invalid_argument: PPL::BD_Shape::generalized_affine_image(e1, r, e2): this->space_dimension() == 2, e2->space_dimension() == 3.

=== test18 === invalid_argument: PPL::BD_Shape::generalized_affine_image(e1, r, e2): this->space_dimension() == 2, e1->space_dimension() == 3. failed tests: test06 FAIL: generalizedaffineimage2

=== test01 === *** bd *** A == 0.5, B >= 5 *** bd.generalized_affine_preimage(B, GREATER_THAN_OR_EQUAL, -1) *** A == 0.5

=== test02 === *** bd *** A == 0.5, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -1) *** false

=== test03 === *** bd *** A == 0.25, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -B+1) *** A == 0.25, B <= -4

=== test04 === *** bd *** A == 0.25, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, B+1) *** A == 0.25, B >= 4, B - A >= 3.75

=== test05 === *** bd *** A >= 0.142857134342193603515625, A <= 0.14285714924335479736328125, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, 2*B+1) *** A >= 0.142857134342193603515625, A <= 0.14285714924335479736328125, B >= 2

=== test06 === *** bd *** A >= 0.199999988079071044921875, A <= 0.20000000298023223876953125, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -2*B+1) *** A >= 0.199999988079071044921875, A <= 0.20000000298023223876953125, B <= -2

=== test07 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 5 Containment does not hold: computed result is A >= 5592405/16777216, A <= 11184811/33554432, B <= -3/2, A - B >= 15379115/8388608 known result is A == 1/3, B <= -3/2, A - B >= 11/6 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, 3*A-2*B+1) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= -1.5, A - B >= 1.83333337306976318359375

=== test08 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 5 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -3*A-2*B+1) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= -2.5

=== test09 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 3.5, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -3*A-2*B+7*C+1, 2) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= 4.66666698455810546875, C <= 2.333333492279052734375, C - A <= 2.0000002384185791015625, B - C <= 2.333333492279052734375

=== test10 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 3.5, C >= 1.39999997615814208984375, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -3*A-2*B-7*C+1, 3) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= -10.1499996185302734375, C >= 1.39999997615814208984375, C <= 2.333333492279052734375, C - A >= 1.06666660308837890625, C - A <= 2.0000002384185791015625

=== test11 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 3.5, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -3*A-2*B+7*C+1, -2) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, C <= 2.333333492279052734375, C - A <= 2.0000002384185791015625

=== test12 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 3.5, C >= 1.39999997615814208984375, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, LESS_THAN_OR_EQUAL, -3*A-2*B-7*C+1, -3) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= -2.91666698455810546875, C >= 1.39999997615814208984375, C <= 2.333333492279052734375, C - A >= 1.06666660308837890625, C - A <= 2.0000002384185791015625

=== test13 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= 3.5, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, GREATER_THAN_OR_EQUAL, -3*A-2*B+7*C+1, -2) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= 11.66666698455810546875, C <= 2.333333492279052734375, C - A <= 2.0000002384185791015625, B - C <= 9.3333339691162109375

=== test14 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= 3.5, C >= 1.39999997615814208984375, C <= 2.333333492279052734375 Containment does not hold: computed result is A >= 5592405/16777216, A <= 11184811/33554432, B <= 2831155/2097152, C >= 11744051/8388608, C <= 4893355/2097152, B - A <= 8528417/8388608, C - A >= 1118481/1048576, C - A <= 8388609/4194304 known result is A == 1/3, B <= 27/20, C >= 7/5, C <= 7/3, B - A <= 61/60, C - A >= 16/15, C - A <= 2, C - B >= 1/20 *** bd.generalized_affine_preimage(B, GREATER_THAN_OR_EQUAL, 3*A-2*B-7*C+1, -3) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B <= 1.349999904632568359375, C >= 1.39999997615814208984375, C <= 2.333333492279052734375, B - A <= 1.01666653156280517578125, C - A >= 1.06666660308837890625, C - A <= 2.0000002384185791015625

=== test15 === *** bd *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, B >= 3.5, C >= 1.39999997615814208984375, C <= 2.333333492279052734375 *** bd.generalized_affine_preimage(B, GREATER_THAN_OR_EQUAL, 3*A-2*B-7*C+1, -3) *** A >= 0.333333313465118408203125, A <= 0.3333333432674407958984375, C >= 1.39999997615814208984375, C <= 2.333333492279052734375, C - A >= 1.06666660308837890625, C - A <= 2.0000002384185791015625 failed tests: test07 test14 FAIL: generalizedaffinepreimage1

Roberto, I didn't have time this morning for the full make check in ppl cvs to run. However I ran it through the float and double sections containing those previously failing tests and they now seem to pass fine on i686-apple-darwin9. I did have to compile ppl cvs with Apple's gcc-4.2 compiler. The current ppl cvs code fails to compile with Apple's stock gcc 4.0.1 compiler. I'll post the error message for that later tonight. Is there a schedule for when ppl cvs will be released as the next version of ppl? Also, have the APIs changed enough between 0.9 and cvs that cloog-ppl from the git won't compile against ppl cvs? I am considering trying to use ppl cvs to build cloog-ppl git and graphite to see if those two failing graphite tests I saw disappear with ppl cvs. Jack

On Thu, Aug 07, 2008 at 08:22:58AM +0200, Roberto Bagnara wrote:

Jack Howarth wrote:

...

I have verified that there are no make check failures in ppl-0.9 on powerpc-apple-darwin9 with Xcode 3.1. The problems appear to be specific to Macintel. I see the same 5 failures when ppl-0.9 is built with Apple's gcc 4.0.1 or gcc 4.2 compilers (so it is unlikely to be a compiler bug). The failing test cases output is appended below.

Jack,

I will check more carefully, but I believe you should not worry about these failures: they are imperfections of the testsuite (which we will fix, if not already fixed in CVS), not of the library. Your attempts to use the CVS HEAD version of the PPL make a lot of sense anyway. Cheers,

Roberto

-- Prof. Roberto Bagnara Computer Science Group Department of Mathematics, University of Parma, Italy http://www.cs.unipr.it/~bagnara/ mailto:bagnara@cs.unipr.it