Changes commited to PARI GIT during the last two weeks

genus2_red5: do full reduction.

This will allow to avoid Tate algorithm later.

1	1	src/basemath/hyperell.c

genus2_eulerfact: return immediately if Euler factor degree is 0

6	3	src/basemath/hyperell.c

polclass: add support for Atkin modular function of level 3

1	0	CHANGES
1	0	src/basemath/polclass.c
47	1	src/basemath/polmodular.c
5	1	src/functions/polynomials/polclass
2	1	src/headers/paripriv.h
290	277	src/test/32/isprime
1	0	src/test/in/polclass

polmodular: handle p2 = 1 correctly

4	3	src/basemath/polmodular.c

polmodular_db_init_allinv: use INV_LAST

2	2	src/basemath/ecpp.c
1	0	src/headers/paripriv.h

test-polclass: fix typo in error message.

2	2	src/test/in/polclass

make test-mf architecture-independent

3	2	src/test/32/mf
3	1	src/test/in/mf

20- mfpetersson() incorrect when one form not cuspidal

1	0	CHANGES
1	1	src/basemath/mftrace.c
2	1	src/test/32/mf
2	0	src/test/in/mf

cleanup ZC_canon, improve ZM_zc_mul_canon_zm

7	11	src/basemath/bibli1.c

preliminary cleanups cvp0_dolll & minim0_dolll

48	74	src/basemath/bibli1.c

split off minim_alloc from cvp_alloc

5	13	src/basemath/bibli1.c

export zv_canon_inplace

1	1	CHANGES
5	0	doc/usersch5.tex
12	0	src/basemath/ZV.c
5	17	src/basemath/qfisom.c
1	0	src/headers/paridecl.h

use pari_APPLY_same

2	7	src/basemath/qfisom.c

17- [libpari] ZC_modhnf, ZM_modhnf

1	0	CHANGES
15	0	doc/usersch6.tex
2	0	src/headers/paridecl.h

mt_err_recover: disable asynchronous cancellation

1	0	src/mt/pthread.c

mt_queue_reset: fix typo in previous commit

2	2	src/mt/pthread.c

aarch64: addllx8/subllx8 use ldp/stp (about 2% faster)

18	26	src/kernel/aarch64/asm0.h

pthread: mt_queue_reset: free pmut, pcond

4	0	src/mt/pthread.c

cleanup theta bench

14	15	src/test/in/theta

strtime: implement days

1	0	CHANGES
6	0	src/language/gplib.c
1	0	src/test/32/str
1	0	src/test/in/str

fix theta bench

6	5	src/test/32/theta
1	1	src/test/in/theta

thetaall_i cleanups

24	18	src/basemath/elltrans.c

unused parameter

4	4	src/basemath/elltrans.c

parforeach([1..1000],j, vector(10^5,i,i)): stack overflow

1	0	CHANGES
1	0	src/language/eval.c

cleanup autojtau

21	25	src/basemath/elltrans.c

move exp(I*Pi*x) to elltrans.c and regenerate paridecl.h

123	0	src/basemath/elltrans.c
0	121	src/basemath/trans3.c
40	38	src/headers/paridecl.h

fix comments

3	1	src/basemath/elltrans.c
1	1	src/basemath/trans3.c

cleanup clearim

17	9	src/basemath/elltrans.c

theta: inline autojz, remove z from autojtau result

24	27	src/basemath/elltrans.c

rewrite mfE2eval

9	6	src/basemath/elltrans.c

set_range: remove the "thickening" by 0.1 when interval is tiny

1	5	src/graph/plotport.c

update bench

3	3	src/test/32/ell
3	2	src/test/in/ell

removed elleisnum optional flag.

If you really want elliptic invariants, use E4/12 and -E6/216.

2	0	CHANGES
2	0	COMPAT
1	1	doc/refcard-ell.tex
1	3	src/basemath/elltrans.c
16	15	src/functions/elliptic_curves/elleisnum
1	1	src/headers/paridecl.h
1	1	src/modules/stark.c

missing part in previous commit (move to BLOCK_SIGINT block)

2	2	src/language/init.c

rename try_to_recover -> try_to_restore + let/unlet it properly

in gp_context_restore: don't call err_printf until it is reset + move
let/unlet inside the BLOCK_SIGINT block.

6	6	src/language/init.c

zeta([2,1+I]) -> error

4	3	src/basemath/trans3.c
2	0	src/test/32/zeta
1	0	src/test/in/zeta

wsigma typo: e need not be a t_REAL

1	1	src/basemath/elltrans.c
7	1	src/test/32/theta
1	0	src/test/in/theta