1 | Hello, |
---|
2 | I'm not sure you'll like all the things I've |
---|
3 | done with your pypar, so any remark you've to |
---|
4 | to, it's welcome. |
---|
5 | |
---|
6 | My problem: I'm trying to do a little master-slave |
---|
7 | parallel program with pypar, something like the |
---|
8 | standard LAM one-step tutorial. |
---|
9 | To do so, I've found that the actual pypar implementation |
---|
10 | lacks of two things (please, please correct me if I'm wrong): |
---|
11 | |
---|
12 | 1. MPI constants in the pypar module (like MPI_ANY_SOURCE, MPI_ANY_TAG) |
---|
13 | 2. The capability to read the MPI_Status structure after |
---|
14 | an MPI_Recv. |
---|
15 | |
---|
16 | for 1. one can always stick the constant -1 inside the code |
---|
17 | but for 2. I've found that the actual implementation doesn't export |
---|
18 | MPI_Status to python. |
---|
19 | |
---|
20 | |
---|
21 | So I've just modified mpi.c (and pypar.py, testpypar.py) to |
---|
22 | 1. export some MPI constants (trivial) |
---|
23 | 2. export MPI_Status structure members along with the error code, in a |
---|
24 | "i(iiii)" (touple of an integer and a four-integer other touple) (ugly) |
---|
25 | 3. modify raw_receive methods accordingly (trivial again, |
---|
26 | but pypar API slighty changes). |
---|
27 | |
---|
28 | |
---|
29 | Even worse, I've modified an old mpi.c version (before bcat), |
---|
30 | so please take into account this. |
---|
31 | |
---|
32 | I've attached all the stuff (along with the ms.py demo program). |
---|
33 | If you want to spend some time reviewing it, I'll be thankful. |
---|
34 | |
---|
35 | Best Regards |
---|
36 | /gp |
---|
37 | |
---|
38 | -- |
---|
39 | "customers, snakes." |
---|
40 | g.p. |
---|
41 | |
---|
42 | "Non è il caldo ma |
---|
43 | sei tu che alzi la temperatura, |
---|
44 | non i fiori ma |
---|
45 | sei tu che profumi l'atmosfera, |
---|
46 | alzati, girati, |
---|
47 | muoviti, risiediti, |
---|
48 | non è il caldo ma |
---|
49 | sei tu che sei bella vera." |
---|
50 | max |
---|
51 | |
---|
52 | "anyone can create a cryptography product that he himself cannot break." |
---|
53 | Bruce Schneier |
---|
54 | |
---|
55 | "If you don't know the things, study them." |
---|
56 | g.p. |
---|
57 | |
---|
58 | "But writing non-free software is another matter. Non-free software is |
---|
59 | a plan for subjugating and dividing the people who use it--an |
---|
60 | antisocial practice. Our goal is to bring that practice to a halt." |
---|
61 | Richard Stallman |
---|
62 | |
---|
63 | Gian Paolo Ciceri via B.Diotti 45 - 20153 Milano MI ITALY |
---|
64 | CTO @ Louise mobile : ++39 348 3658272 |
---|
65 | eMail : gp.ciceri@acm.org, |
---|
66 | gp.ciceri@louise.it |
---|
67 | webSite: http://www.louise.it |
---|
68 | ICQ : 94620118 |
---|