September 2004 - Darwin-dev

Re: Using dead notification with mach ports
by Jim Magee 10 Sep '04

10 Sep '04

On Sep 10, 2004, at 12:47 PM, Stiphane Letz wrote: Yes. Yes, but now requesting a MACH_NOTIFY_NO_SENDERS notification. So the system would be : 3) client do a RPC call to register itself : - server add the new private port in its portset - server request a "no-senders" notification on the new private port 5) client uses its private port to do RPC calls to the server _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored. I was thinking that doing the "task_get_bootstrap_port" and "bootstrap_look_up(fBootPort, "server", &fServerPort)" code on the client side (each client does that) was actually creating a new port for each client, but it seems this is not the case? All clients see the same "server" port allocated by the server? Is that correct? SInce the name "port" is sometimes interchangeably used to talk about the end-points of the communication channel and to talk about the channel itself (i.e. the message queue), I can understand your confusion. I will try to use "message queue" when I mean that. There is only one Mach message queue in this case. That's the one the server creates and then copies a send right to it into the bootstrap namespace (via the bootstrap_register() RPC). When each client does a bootstrap_look_up() RPC, they are each copied a send right to that single Mach message queue (as part of the reply to the RPC). The thing is that doing this way the server was able to handle RPC call from all clients in the "mach_msg_server" call..... And that can continue regardless of whether there is a single message queue (directly passed to the mach_msg_server() call) or multiple message queues added to a portset and then that passed to the mach_msg_server() call instead. When you say "the server creates a new, private, Mach port for the client and adds it to the portset that it waits on", do you mean the server will create a additional port for each client and this additional port can be used also to make RPC calls from the client to the server? Then when you say "The server can then request a "no-senders" notification " do it means that the server has to use the "mach_port_request_notification" fucntion (not the client as I was doing..) 1) server first create the "server" port (bootstrap_register(fBootPort, "server", fServerPort)...) A kind of general entry point. 2) client get this port (task_get_bootstrap_port and bootstrap_look_up(fBootPort, "server", &fServerPort) to get this port to communicate with the server - server receives this RPC call and create a new private port for the client 4) client get this new private port from the server (same method as 2 ) This is where you went astray. The clients wouldn't look the new message queue up in some global namespace, but instead it would be returned to them (privately) as part of the reply to this "connection request" RPC. 6) if client goes away, they the server receives a notification on the client private port and thus know which client went away Is this the idea?

1 0

Calling shm_open from a CFM app.
by Corey O'Connor 10 Sep '04

10 Sep '04

I'm having trouble calling shm_open from a CFM app. I've successfully called other MachO functions from this application by using CFBundleGetFunctionPointerForName. When I call shm_open, however, it seems to be stomping on memory. So the call fails and shortly afterwards my app crashes. I suspect it may have to do with shm_open being variadic. Is there any special magic I need to perform to call a variadic function? Or do I have to write a wrapper framework to call this function? I have tested the same call outside of the CFM application with success. Thanks! -- -Corey O'Connor _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Re: Using dead notification with mach ports
by Stéphane Letz 10 Sep '04

10 Sep '04

The correct answer almost never includes task_for_pid(). ;-) So the system would be : - server add the new private port in its portset - server request a "no-senders" notification on the new private port 4) client get this new private port from the server (same method as 2 ) 5) client uses its private port to do RPC calls to the server Is this the idea? Thanks Stephane Letz _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored. Besides, the task_for_pid() call you make above is just getting your own (client) task port. So, server_port should just be the same as what mach_task_self() returns. You didn't need to call task_for-pid() to get that. The second problem is that you are trying to use a send right (the client's reference to fServerPort) to create a send-once right (the one you try to make as part of the notification request). You simply cannot create a send-once right from a send right. Only the receive right can be used to make a send-once right. And that means that only the server itself can make this request. In trying to fix this, it is important to think about notifications in general. All Mach port notification messages are "port namespace relative" (only the name of the port is sent in the message). Within each task's port namespace, the mach_task_self() port likely has the same low-number name (because it is exposed during the pre-main() initialization of the process). So, nearly every one of the notification messages received by the server would look identical. Nothing would actually identify the client in any way known to the server. What you need is a reference, held by the server, that identifies each client. Then the server can request a notification about that reference. It is tempting to just have each client send the server their own task port to identify themselves. But that's a security issue if you don't trust the server whole-heartedly. You could just send a PID and have the server do the task_for_pid() itself. But then the client may just be spoofing the PID. In Panther 10.3.4 and above you could get a "trusted path" PID from the Mach message audit trailer to get around the spoofing. But this is all too kludgy. I f the server really does maintain state for each client independently, it should probably maintain a private connection with each such client. Just use the fServerPort (the one found in the bootstrap namespace) to have each client make a private connection to the server (the server creates a new, private, Mach port for the client and adds it to the portset that it waits on). The server can then request a "no-senders" notification on that private connection port to determine when the client goes away. --Jim I was thinking that doing the "task_get_bootstrap_port" and "bootstrap_look_up(fBootPort, "server", &fServerPort)" code on the client side (each client does that) was actually creating a new port for each client, but it seems this is not the case? All clients see the same "server" port allocated by the server? Is that correct? The thing is that doing this way the server was able to handle RPC call from all clients in the "mach_msg_server" call..... When you say "the server creates a new, private, Mach port for the client and adds it to the portset that it waits on", do you mean the server will create a additional port for each client and this additional port can be used also to make RPC calls from the client to the server? Then when you say "The server can then request a "no-senders" notification " do it means that the server has to use the "mach_port_request_notification" fucntion (not the client as I was doing..) 1) server first create the "server" port (bootstrap_register(fBootPort, "server", fServerPort)...) A kind of general entry point. 2) client get this port (task_get_bootstrap_port and bootstrap_look_up(fBootPort, "server", &fServerPort) to get this port to communicate with the server 3) client do a RPC call to register itself : - server receives this RPC call and create a new private port for the client 6) if client goes away, they the server receives a notification on the client private port and thus know which client went away

1 0

Re: Using dead notification with mach ports
by Jim Magee 10 Sep '04

10 Sep '04

On Sep 10, 2004, at 6:40 AM, Stiphane Letz wrote: Hi I'm developing a server/client system and would like to use the MACH_NOTIFY_DEAD_NAME system so that the server can know when clients disappear abnormally (crash...) ...[snip]... Then I'm trying to add a dead notification to notify the server when clients disappear : On the client side : mach_port_t server_port; mach_port_t old_port; if (task_for_pid(mach_task_self(),getpid(),&server_port) != KERN_SUCCESS) { printf("NotifyDead: can't find mach task_for_pid"); return false; } kern_return_t result; if ((result = mach_port_request_notification(mach_task_self(),server_port, MACH_NOTIFY_DEAD_NAME, 0,fServerPort,MACH_MSG_TYPE_MAKE_SEND_ONCE, &old_port)) != KERN_SUCCESS) { printf("NotifyDead %s\n",mach_error_string(result)); return false; } But I always get a "NotifyDead (ipc/send) invalid port right" error message. Any idea of what is wrong? Is is the right way to do? Any link to code/examples that implements this correctly? The correct answer almost never includes task_for_pid(). ;-) --Jim Thanks _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored. Besides, the task_for_pid() call you make above is just getting your own (client) task port. So, server_port should just be the same as what mach_task_self() returns. You didn't need to call task_for-pid() to get that. The second problem is that you are trying to use a send right (the client's reference to fServerPort) to create a send-once right (the one you try to make as part of the notification request). You simply cannot create a send-once right from a send right. Only the receive right can be used to make a send-once right. And that means that only the server itself can make this request. In trying to fix this, it is important to think about notifications in general. All Mach port notification messages are "port namespace relative" (only the name of the port is sent in the message). Within each task's port namespace, the mach_task_self() port likely has the same low-number name (because it is exposed during the pre-main() initialization of the process). So, nearly every one of the notification messages received by the server would look identical. Nothing would actually identify the client in any way known to the server. What you need is a reference, held by the server, that identifies each client. Then the server can request a notification about that reference. It is tempting to just have each client send the server their own task port to identify themselves. But that's a security issue if you don't trust the server whole-heartedly. You could just send a PID and have the server do the task_for_pid() itself. But then the client may just be spoofing the PID. In Panther 10.3.4 and above you could get a "trusted path" PID from the Mach message audit trailer to get around the spoofing. But this is all too kludgy. I f the server really does maintain state for each client independently, it should probably maintain a private connection with each such client. Just use the fServerPort (the one found in the bootstrap namespace) to have each client make a private connection to the server (the server creates a new, private, Mach port for the client and adds it to the portset that it waits on). The server can then request a "no-senders" notification on that private connection port to determine when the client goes away. Stephane Letz _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Re: API to open the firewall?
by Joshua Graessley 10 Sep '04

10 Sep '04

There is no API for doing this. There are ways to do this using ioctls to configure the firewall, but this is error prone. It is unlikely there will ever be a straightforward API for simply opening up the ports. If the user enabled the firewall, it is usually because they don't trust the software they're running, they were influenced by marketing "firewall makes it safer", or they've been reading about the miseries of another platform out there. At any rate, if there was an API for opening ports (bypassing the firewall), it would pretty much eliminate the functionality of the firewall. I believe there is a bug filed about giving third parties a way to add ports to the list in the firewall preferences pane. This is slightly off topic for darwin-dev. There is a mac network programming list that might be a better place. If you are an ADC member, DTS might also be able to help. There are better solutions, such as a system wide way of controlling which users and which processes have permission to open sockets and listen for inbound connections. -josh On Sep 9, 2004, at 11:28 PM, Braren, Arfst RD-PN34 wrote: > Hi all, > An client/server application is using some ports to talk http (630x > and 5200x depends on whats available). When firewall is enabled the > client can't talk with the server - the user has to modify the > firewall preferences. - not so easy because we are using port ranges. > > Is there an API that allows an app to register ports at the firewall? > Is there a known way for the Installer to do so? > > Thanks > A.Braren > _______________________________________________ > darwin-development mailing list | darwin-development(a)lists.apple.com > Help/Unsubscribe/Archives: > http://www.lists.apple.com/mailman/listinfo/darwin-development > Do not post admin requests to the list. They will be ignored. [demime 0.98b removed an attachment of type application/pkcs7-signature which had a name of smime.p7s] _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Using dead notification with mach ports
by Stéphane Letz 10 Sep '04

10 Sep '04

Hi I'm developing a server/client system and would like to use the MACH_NOTIFY_DEAD_NAME system so that the server can know when clients disappear abnormally (crash...) On the server side : Registration of the server port: ...... ...... mach_port_t task = mach_task_self(); if (task_get_bootstrap_port(task, &fBootPort) != KERN_SUCCESS) { printf("AllocatePort: Can't find bootstrap mach port"); return false; } if (mach_port_allocate(task, MACH_PORT_RIGHT_RECEIVE, &fServerPort) != KERN_SUCCESS) { printf("AllocatePort: can't allocate mach port"); return false; } if (mach_port_insert_right(task, fServerPort, fServerPort, MACH_MSG_TYPE_MAKE_SEND) != KERN_SUCCESS) { printf("AllocatePort: error inserting mach rights"); return false; } if (bootstrap_register(fBootPort, "server", fServerPort) != KERN_SUCCESS) { printf("Allocate: can't check in mach port"); return false; } ...... ...... Then in a thread to get messages coming from clients: boolean_t server_demux(mach_msg_header_t *Request,mach_msg_header_t *Reply) { printf("received message on port %x\n", Request->msgh_local_port); ..... ..... return TRUE; } kern_return_t res = mach_msg_server(server_demux,1024,fServerPort,0); On the client side : Connection to the server port: ...... ...... mach_port_t task = mach_task_self(); if (task_get_bootstrap_port(task, &fBootPort) != KERN_SUCCESS) { printf("ConnectPort: can't find bootstrap port"); return false; } if (bootstrap_look_up(fBootPort, "server", &fServerPort) != KERN_SUCCESS) { printf("ConnectPort: can't find mach server port"); return false; } ...... ...... This works correctly to use MIG generated RPC between the clients and the server. Then I'm trying to add a dead notification to notify the server when clients disappear : On the client side : mach_port_t server_port; mach_port_t old_port; if (task_for_pid(mach_task_self(),getpid(),&server_port) != KERN_SUCCESS) { printf("NotifyDead: can't find mach task_for_pid"); return false; } kern_return_t result; if ((result = mach_port_request_notification(mach_task_self(),server_port, MACH_NOTIFY_DEAD_NAME, 0,fServerPort,MACH_MSG_TYPE_MAKE_SEND_ONCE, &old_port)) != KERN_SUCCESS) { printf("NotifyDead %s\n",mach_error_string(result)); return false; } But I always get a "NotifyDead (ipc/send) invalid port right" error message. Any idea of what is wrong? Is is the right way to do? Any link to code/examples that implements this correctly? Thanks Stephane Letz _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

API to open the firewall?
by Braren, Arfst RD-PN34 10 Sep '04

10 Sep '04

Hi all, An client/server application is using some ports to talk http (630x and 5200x depends on whats available). When firewall is enabled the client can't talk with the server - the user has to modify the firewall preferences. - not so easy because we are using port ranges. Is there an API that allows an app to register ports at the firewall? Is there a known way for the Installer to do so? Thanks A.Braren _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Re: trouble with valarray and gcc 3.3
by Jerry Pendergraft 10 Sep '04

10 Sep '04

err - maybe a cut and paste kinda error but line 2 of the code example: int main{ changed to: int main(){ seems to fix the problem? -- Jerry Pendergraft jerry.pendergraft(a)endocardial.com Endocardial Solutions voice: 651-523-6935 1350 Energy Lane, Suite 110 fax: 651-644-7897 St Paul, MN 55108-5254 mobile: 651-491-0163 On Thu, 9 Sep 2004, Doug McCoy wrote: > On Sep 9, 2004, at 9:58 PM, Joseph Kim wrote: > > > thanks for the quick reply. my bad on that. but the problem still exists. > > ----code---- > #include <valarray> > int main{ > std::valarray<int> v1; // construct an empty valarray > std::valarray<int> v2(1,3); // construct a valarray of three > // elements, all initialized to 1 > v1.resize(3,2); // resize the first valarray to three > // elements, all initialized to 2 > std::valarray<int> v3(v1); // v3 gets a copy of v1's elements. > } > ---end--- > > and > > ---errors--- > test.cpp:2: error: syntax error before `{' token > test.cpp:6: error: syntax error before `.' token > test.cpp:8: error: `v1' was not declared in this scope > test.cpp:9: error: parse error before `}' token > ---end--- > > still having trouble. > > ??? > > > > __________________________________ > Do you Yahoo!? > Yahoo! Mail is new and improved - Check it out! > http://promotions.yahoo.com/new_mail > _______________________________________________ > darwin-development mailing list | darwin-development(a)lists.apple.com > Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development > Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Re: trouble with valarray and gcc 3.3
by Doug McCoy 10 Sep '04

10 Sep '04

On Sep 9, 2004, at 9:58 PM, Joseph Kim wrote: >>All other lines are fine as global variable declarations/initializations. >>The offending line doesn't belong at this scope. >> >>I.e.: >> >>#include <valarray> >> >> >>int main() >>{ >> // ... insert your code here. >> // ... the offending line is fine as a statement within a function. >> >> >> return 0; >>} >> >>Not an issue with Darwin GCC. >> >>As an aside, I always use "c++" rather than "gcc" to drive the linker to make >>sure all the correct libraries and object files are brought in if at least >>one of the source files is a C++ source file (too lazy to try to supply all the >>right linker input manually :) thanks for the quick reply. my bad on that. but the problem still exists. ----code---- #include <valarray> int main{ std::valarray<int> v1; // construct an empty valarray std::valarray<int> v2(1,3); // construct a valarray of three // elements, all initialized to 1 v1.resize(3,2); // resize the first valarray to three // elements, all initialized to 2 std::valarray<int> v3(v1); // v3 gets a copy of v1's elements. } ---end--- and ---errors--- test.cpp:2: error: syntax error before `{' token test.cpp:6: error: syntax error before `.' token test.cpp:8: error: `v1' was not declared in this scope test.cpp:9: error: parse error before `}' token ---end--- still having trouble. ??? __________________________________ Do you Yahoo!? Yahoo! Mail is new and improved - Check it out! http://promotions.yahoo.com/new_mail _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored.

1 0

Re: trouble with valarray and gcc 3.3
by Joseph Kim 10 Sep '04

10 Sep '04

On Sep 9, 2004, at 9:17 PM, Doug McCoy wrote: -----start---- #include <valarray> std::valarray<int> v1; // construct an empty valarray std::valarray<int> v2(1,3); // construct a valarray of three // elements, all initialized to 1 v1.resize(3,2); // resize the first valarray to three ^^^^^^^^^^^^^^^^ // elements, all initialized to 2 std::valarray<int> v3(v1); // v3 gets a copy of v1's elements. -----end----- I.e.: #include <valarray> int main() { // ... insert your code here. // ... the offending line is fine as a statement within a function. return 0; } Not an issue with Darwin GCC. 10Q. Joe _______________________________________________ darwin-development mailing list | darwin-development(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-development Do not post admin requests to the list. They will be ignored. All other lines are fine as global variable declarations/initializations. The offending line doesn't belong at this scope. As an aside, I always use "c++" rather than "gcc" to drive the linker to make sure all the correct libraries and object files are brought in if at least one of the source files is a C++ source file (too lazy to try to supply all the right linker input manually :)

1 0