mquinson [Tue, 6 Jul 2004 01:50:03 +0000 (01:50 +0000)]
More DEBUG call
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@177
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 6 Jul 2004 01:49:31 +0000 (01:49 +0000)]
Eradicate fprintf, use proper logging function
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@176
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 6 Jul 2004 01:22:43 +0000 (01:22 +0000)]
gras_trp_socket_new is in charge of pushing it into the set of known sockets
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@175
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 6 Jul 2004 01:01:17 +0000 (01:01 +0000)]
Make sure the process get the --gras-log argument in SG mode (where the arguments are usually given by the deployment file)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@174
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 6 Jul 2004 00:43:17 +0000 (00:43 +0000)]
Each process have to run gras_init/exit (one day, they'll have separate log settings)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@173
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 6 Jul 2004 00:31:24 +0000 (00:31 +0000)]
Let it work in SG (gras_init and gras_exit called only when needed)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@172
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:53:46 +0000 (23:53 +0000)]
Some more code factorisation (registering the new socket into the set of known sockets is done by trp_socket_new instead of all over the place)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@171
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:34:03 +0000 (23:34 +0000)]
free the callback list on gras_process_exit
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@170
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:21:09 +0000 (23:21 +0000)]
move the dynar of all known sockets from RL-only to procdata (SG-wanna-accept change)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@169
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:19:44 +0000 (23:19 +0000)]
sockets is now part of the process data
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@168
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:18:02 +0000 (23:18 +0000)]
use gras_trp_socket_new when accepting instead of mallocing ourself to get all the good initialization from there
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@167
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:15:00 +0000 (23:15 +0000)]
initialize the buffer before use, even if the part actually used was already initialized. This prevent valgrind to spit an error about data passed to syscall not being [completely] initialized
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@166
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:12:54 +0000 (23:12 +0000)]
add a valgrind target to the makefile, but not for the dict_crash since it takes for ever if we do run it under valgrind
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@165
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 23:11:31 +0000 (23:11 +0000)]
process_{init,exit} from within gras_{init,exit}
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@164
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 22:31:17 +0000 (22:31 +0000)]
Plug memleaks on exit (I know, it's useless)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@163
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 22:27:26 +0000 (22:27 +0000)]
Plug a memleak on the name of the incomming messages
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@162
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 22:18:34 +0000 (22:18 +0000)]
Plug a memleak on sock_close (on peer_name)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@161
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 5 Jul 2004 22:15:54 +0000 (22:15 +0000)]
Initialize whether the accepted sock is raw or not from what the server socket is.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@160
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 23:30:56 +0000 (23:30 +0000)]
revert last change trying to inject the \0 after the name, since it was mainly cosmetic and break things around
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@159
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 22:39:44 +0000 (22:39 +0000)]
When inserting into the dict, save the trailing '\0' (ie, use strlen+1 as size)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@158
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 20:15:53 +0000 (20:15 +0000)]
kill dead code
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@157
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:49:21 +0000 (07:49 +0000)]
Make sure that examples/ping/test_rl and examples/ping/test_sg are executable
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@156
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:48:52 +0000 (07:48 +0000)]
dummy scripts passing the args to the programs. Damn automake limitation
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@155
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:46:26 +0000 (07:46 +0000)]
Let it work in RL (yuhu)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@154
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:43:43 +0000 (07:43 +0000)]
Propagate the interface cleanup from last week in datadesc, ie remove a
superfluous level of indirection. User pass adress of variable containing
data (both when sending and receiving), and not of a variable being a
pointer to the data. Let's say that I like it better ;)
The price for that is constructs like "int msg=*(int*)payload" in handlers,
but it's a fine price, IMHO.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@153
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:41:19 +0000 (07:41 +0000)]
new function: gras_datadesc_size to allow messaging layer to malloc the needed space
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@152
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:32:43 +0000 (07:32 +0000)]
adapt to the new syntax of socket_close
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@151
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:31:12 +0000 (07:31 +0000)]
new function: gras_datadesc_size to allow messaging layer to malloc the needed space
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@150
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:21:25 +0000 (07:21 +0000)]
gras_socket_close now expect the socket to close (and not its address to
put NULL in it after it).
This is because the socket passed to handlers is one of their argument (=>
not writable).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@149
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 07:17:45 +0000 (07:17 +0000)]
-=gs_example ; +=datadesc_usage
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@148
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 04:14:06 +0000 (04:14 +0000)]
Fix the ugly case of reference to dynamic array.
The size of this array comes from the network.
It is needed both at the reference reception level (to allocate enough
space) and at the array reception level (to fill enough room).
It is then passed as an argument of the recursive function. This is a crude
hack, but I was told that working code is sometimes better than neat one ;)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@147
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 04:10:02 +0000 (04:10 +0000)]
Plug a tiny leak
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@146
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Sat, 3 Jul 2004 03:08:32 +0000 (03:08 +0000)]
Create a new log channel tbx containing dict, set, log, dynar, config (to shut
them all up in one shot)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@145
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 29 Jun 2004 05:09:30 +0000 (05:09 +0000)]
Make it easy to specify the version of automake to use (it's 1.7 here on graal)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@144
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 29 Jun 2004 05:07:17 +0000 (05:07 +0000)]
New configure (generated on Graal since my hard disk died, and since the installed version here in the lab is automake-1.4, erk)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@143
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Tue, 29 Jun 2004 04:06:31 +0000 (04:06 +0000)]
Cruft eradication
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@142
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:55:34 +0000 (17:55 +0000)]
document last changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@141
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:54:53 +0000 (17:54 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@140
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:54:14 +0000 (17:54 +0000)]
Change the default port
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@139
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:52:38 +0000 (17:52 +0000)]
More work on SG, almost done
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@138
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:43:38 +0000 (17:43 +0000)]
Give it the possibility to do only writes or only reads (so that I can work on portability); make it work, even with cycles (beside with strings)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@137
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:42:04 +0000 (17:42 +0000)]
make a gras_msg_recv_no_malloc function to make clear that if you don't NULL its argument, you'll get into trouble (interface simplification again)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@136
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:41:42 +0000 (17:41 +0000)]
register the message even if it's already there with the same payload so that the old payload gets free'd when its refcount becomes zero; make a gras_msg_recv_no_malloc function to make clear that if you don't NULL its argument, you'll get into trouble (interface simplification again)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@135
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:37:19 +0000 (17:37 +0000)]
remove one indirection of gras_datadesc_recv to ease its semantic; plug a few leaks; make gras_datadesc_type_cmp do its job
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@134
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:35:12 +0000 (17:35 +0000)]
remove one indirection of gras_datadesc_recv to ease its semantic
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@133
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:14:18 +0000 (17:14 +0000)]
backtraces are useless
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@132
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:14:03 +0000 (17:14 +0000)]
prototypes of some new debugging functions
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@131
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:13:27 +0000 (17:13 +0000)]
Cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@130
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:13:05 +0000 (17:13 +0000)]
implement gras_procdata_exit to plug the leaks
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@129
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:12:35 +0000 (17:12 +0000)]
implement gras_process_exit (using gras_procdata_exit) to plug the leaks
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@128
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:11:37 +0000 (17:11 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@127
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:10:56 +0000 (17:10 +0000)]
plug a stupid leak
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@126
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:08:48 +0000 (17:08 +0000)]
Use refcounters on datadescription types so that messages can have pointer to datadescription directly (and not their name), and still deal with multiple creations of the types (in SG, each process create them) and without leaking on them
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@125
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:05:37 +0000 (17:05 +0000)]
New (debugging) function: gras_datadesc_arch_name
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@124
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:04:47 +0000 (17:04 +0000)]
New (debugging) functions: gras_datadesc_get_{name,id} gras_datadesc_type_dump
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@123
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:03:26 +0000 (17:03 +0000)]
More verbose
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@122
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:03:08 +0000 (17:03 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@121
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 21 Jun 2004 17:02:45 +0000 (17:02 +0000)]
Plug a memleak when using gras_dynar_cursor_rm
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@120
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:20:18 +0000 (18:20 +0000)]
Initialize GRAS before use ;)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@119
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:12:42 +0000 (18:12 +0000)]
automatic update
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@118
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:10:50 +0000 (18:10 +0000)]
porting to new standards
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@117
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:08:17 +0000 (18:08 +0000)]
Add an implementation trick to speed up SG realm
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@116
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:05:05 +0000 (18:05 +0000)]
Various cleanups; SG almost work
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@115
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:04:03 +0000 (18:04 +0000)]
More work on this. Ok, it does not really work yet, but it compiles ;)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@114
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:02:51 +0000 (18:02 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@113
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:01:53 +0000 (18:01 +0000)]
Handle the fact that the RL plugin do not exist in SG mode (and vice versa); plugin names are lowcase; more debug message
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@112
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 18:00:00 +0000 (18:00 +0000)]
Implement select over SG (badly since there is so few support within SG); Add stupid implementation of the RL plugins to use when in SG mode
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@111
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 17:57:46 +0000 (17:57 +0000)]
Add stupid implementation of the SG plugins to use when in RL mode
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@110
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 17:53:49 +0000 (17:53 +0000)]
Eradicate two small, nasty and stupid bugs
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@109
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 17:51:10 +0000 (17:51 +0000)]
A debug msg more
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@108
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 18 Jun 2004 17:47:36 +0000 (17:47 +0000)]
better split SG/RL/common ; remove old cruft
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@107
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:28:06 +0000 (20:28 +0000)]
Continuation of the internal cleanup; Copy/paste of the SG implementation (not working yet, but soon)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@106
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:25:59 +0000 (20:25 +0000)]
Field renaming in the plugin struct to cleanup; raw is now a field of the socket, not a creation function argument
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@105
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:25:01 +0000 (20:25 +0000)]
Got all of this in the new generation GRAS
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@104
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:24:30 +0000 (20:24 +0000)]
The core is in charge of mallocing for the transport plugins (instead of the plugin creation function themselves); raw is now a field of the socket, not a creation function argument
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@103
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:22:52 +0000 (20:22 +0000)]
get/set instead of retrieve/insert
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@102
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 14 Jun 2004 20:21:17 +0000 (20:21 +0000)]
More to do
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@101
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 11 Jun 2004 21:09:06 +0000 (21:09 +0000)]
+=datadesc_usage
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@100
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 11 Jun 2004 20:34:48 +0000 (20:34 +0000)]
- Reput hook for raw sockets, needed for BW experiments
- kill a few lines of dead code
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@99
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 11 Jun 2004 20:29:19 +0000 (20:29 +0000)]
Remove a unused variable
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@98
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 11 Jun 2004 19:39:22 +0000 (19:39 +0000)]
[Data description]
- Interface cleanup.
. gras_datadesc_by_name returns the searched type or NULL.
That way, no variable is needed to use a type desc once, which makes
the code clearer.
. gras_datadesc_declare_[struct|union]_append_name is removed. The last
two parameters were strings (field name, type name), leading to
common errors.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@97
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 11 Jun 2004 19:12:50 +0000 (19:12 +0000)]
More (forgotten) s/retrive/get/ ; s/insert/set/
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@96
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 4 Jun 2004 22:45:11 +0000 (22:45 +0000)]
Update of external scripts
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@95
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:24:48 +0000 (22:24 +0000)]
API cleanup: s/dict_insert/dict_set/ and s/dict_retrieve/dict_get/ for consistency with dynar
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@94
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:24:19 +0000 (22:24 +0000)]
new version of a script not being mine
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@93
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:16:04 +0000 (22:16 +0000)]
API cleanup: s/dict_insert/dict_set/ and s/dict_retrieve/dict_get/ for consistency with dynar
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@92
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:15:13 +0000 (22:15 +0000)]
redo the instruction for the error case (please mail the results to...)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@91
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:04:55 +0000 (22:04 +0000)]
fill the tree before manually removing elements form it + cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@90
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 3 Jun 2004 22:04:10 +0000 (22:04 +0000)]
Do not assert out when the user tries to remove data from a NULL dict: that's a noop ;)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@89
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 28 Apr 2004 11:55:39 +0000 (11:55 +0000)]
Document last changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@88
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 28 Apr 2004 11:55:02 +0000 (11:55 +0000)]
gs est parti
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@87
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 26 Apr 2004 10:15:11 +0000 (10:15 +0000)]
Work to port on new API, to be terminated
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@86
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 26 Apr 2004 10:14:42 +0000 (10:14 +0000)]
Generate code compatible with the new interfaces
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@85
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 26 Apr 2004 10:14:05 +0000 (10:14 +0000)]
Update to reflect last code changes
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@84
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Mon, 26 Apr 2004 10:11:28 +0000 (10:11 +0000)]
[Messaging]
- Porting to new standards.
[Data description]
- interface cleanup.
There is no bag anymore, no need to take extra provision to mask the
pointers behind "ID".
Better splitup of functions between files create/exchange/convert.
This is still a bit artificial since convert and receive are so
interleaved, but anyway.
[Virtu(process)]
- add a queued message list to procdata (the ones not matching criteria
in msg_wait)
- factorize some more code between SG and RL wrt procdata
[Tests]
- use gras_exit in example to track memleaks
- get rid of gs_example now that GS is properly integrated into gras
- update run_test to integrate the lastest tests (datadesc)
[Logging]
- rename WARNINGn macros to WARNn since it prooved error-prone
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@83
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Wed, 21 Apr 2004 18:44:57 +0000 (18:44 +0000)]
[Data description]
- register init/exit functions within gras module mecanism
- send/receive function.
Convertion is not implemented, but short-cutted if not needed.
struct/array elements are sent one by one (instead of block-wise), but
nobody really cares (yet). Get a prototype before optimizing.
- tests (using a file socket) for DD send/receive on:
- base types: int, float
- array: fixed size, string (ie ref to dynamic string)
- structure: homogeneous, heterogeneous
- chained list, graph with cycle
Believe it or not, valgrind is not too unhappy with the results. The
cycle happily segfaults, but the others are ok. And I'm sick of pointers
for now.
[Transport]
[File plugin]
- Bugfix when using a filename explicitely (instead of '-')
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@82
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Thu, 15 Apr 2004 00:17:02 +0000 (00:17 +0000)]
2004-04-09 Martin Quinson <martin.quinson@tuxfamily.org>
[Transport plugins]
- factorize more code between RL and SG in socket creation
- Complete the implementation and tests of:
o TCP
o file (only in RL, and mainly for debugging)
I lost 3 days to design a portable address resolver, and then decided
that the prototype mainly have to run on my box.
Addressing portability too early may be like optimizing too early :-/
[Tests]
- use gras_init in the Tests instead of the crappy parse_log_opt
(the latter function is removed)
[Conditional execution]
- New functions: gras_if_RL/gras_if_SG (basic support for this)
[Code reorganisation]
- Get rid of libgrasutils.a since it makes more trouble than it solves.
Build examples against the RL library, since there is no way to disable
its creation for now.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@81
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 9 Apr 2004 18:31:01 +0000 (18:31 +0000)]
cosmetics
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@80
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 9 Apr 2004 18:29:44 +0000 (18:29 +0000)]
rename bloc_send/recv to chunk_send/recv because I had a doubt on bloc meaning in english
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@79
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
mquinson [Fri, 9 Apr 2004 18:28:19 +0000 (18:28 +0000)]
Redirect chunk_send/recv functions to the plugin version
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/simgrid/simgrid/trunk@78
48e7efb5-ca39-0410-a469-
dd3cf9ba447f
