* [Xenomai-help] IRQ1 forwarded to Linux? @ 2006-04-27 11:17 ROSSIER Daniel 2006-04-27 11:57 ` Philippe Gerum 2006-04-27 11:59 ` Philippe Gerum 0 siblings, 2 replies; 6+ messages in thread From: ROSSIER Daniel @ 2006-04-27 11:17 UTC (permalink / raw) To: xenomai [-- Attachment #1: Type: text/plain, Size: 653 bytes --] Hello, We have a small Xeno application which implements an interrupt service (ISR) on IRQ 1 (keyboard) - quite stupid, but just for testing. If we understand well, returning from the ISR with RET_INTR_HANDLED should not propagate the IRQ to the other domains, therefore to the Linux one, right? Why can we then keep working with the shell? (I expect that we can not interact with the shell, even not see the echoed char.). Does Xenomai patch the keyb driver? (We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on the /proc/interrupts, as expected). Thanks for your help Daniel [-- Attachment #2: Type: text/html, Size: 3662 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Xenomai-help] IRQ1 forwarded to Linux? 2006-04-27 11:17 [Xenomai-help] IRQ1 forwarded to Linux? ROSSIER Daniel @ 2006-04-27 11:57 ` Philippe Gerum 2006-04-27 11:59 ` Philippe Gerum 1 sibling, 0 replies; 6+ messages in thread From: Philippe Gerum @ 2006-04-27 11:57 UTC (permalink / raw) To: ROSSIER Daniel; +Cc: xenomai ROSSIER Daniel wrote: > Hello, > > > > We have a small Xeno application which implements an interrupt service > (ISR) on IRQ 1 (keyboard) – quite stupid, but > What does the following say? $ cat /proc/ipipe/Linux > just for testing. If we understand well, returning from the ISR with > RET_INTR_HANDLED should not propagate the > > IRQ to the other domains, therefore to the Linux one, right? > > Why can we then keep working with the shell? (I expect that we can not > interact with the shell, even not see the echoed char.). > > > > Does Xenomai patch the keyb driver? > > > > (We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > the /proc/interrupts, as expected). > > > > Thanks for your help > > > > Daniel > > > ------------------------------------------------------------------------ > > _______________________________________________ > Xenomai-help mailing list > Xenomai-help@domain.hid > https://mail.gna.org/listinfo/xenomai-help -- Philippe. ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Xenomai-help] IRQ1 forwarded to Linux? 2006-04-27 11:17 [Xenomai-help] IRQ1 forwarded to Linux? ROSSIER Daniel 2006-04-27 11:57 ` Philippe Gerum @ 2006-04-27 11:59 ` Philippe Gerum 1 sibling, 0 replies; 6+ messages in thread From: Philippe Gerum @ 2006-04-27 11:59 UTC (permalink / raw) To: ROSSIER Daniel; +Cc: xenomai ROSSIER Daniel wrote: > Hello, > > > > We have a small Xeno application which implements an interrupt service > (ISR) on IRQ 1 (keyboard) – quite stupid, but > Sorry, - $ cat /proc/ipipe/Linux + $ cat /proc/ipipe/Xenomai when your test is loaded would be useful to understand why Linux still gets the kbd IRQs. > just for testing. If we understand well, returning from the ISR with > RET_INTR_HANDLED should not propagate the > > IRQ to the other domains, therefore to the Linux one, right? > > Why can we then keep working with the shell? (I expect that we can not > interact with the shell, even not see the echoed char.). > > > > Does Xenomai patch the keyb driver? > No. > > > (We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > the /proc/interrupts, as expected). > > > > Thanks for your help > > > > Daniel > > > ------------------------------------------------------------------------ > > _______________________________________________ > Xenomai-help mailing list > Xenomai-help@domain.hid > https://mail.gna.org/listinfo/xenomai-help -- Philippe. ^ permalink raw reply [flat|nested] 6+ messages in thread
[parent not found: <1146143577.3671.3.camel@domain.hid>]
* Re: Re: [Xenomai-help] IRQ1 forwarded to Linux? [not found] <1146143577.3671.3.camel@domain.hid> @ 2006-04-27 14:44 ` Guillaume Boutillier 2006-04-27 14:56 ` Philippe Gerum 0 siblings, 1 reply; 6+ messages in thread From: Guillaume Boutillier @ 2006-04-27 14:44 UTC (permalink / raw) To: xenomai Hello, > > ROSSIER Daniel wrote: > > > Hello, > > > > > > > > > > > > We have a small Xeno application which implements an interrupt service > > > (ISR) on IRQ 1 (keyboard) – quite stupid, but > > > > > > > Sorry, > > > > - $ cat /proc/ipipe/Linux > > + $ cat /proc/ipipe/Xenomai I checked when I run my application... But I couldn't understand the meaning of the different flags : accepted, passed and grabbed?! > > > > when your test is loaded would be useful to understand why Linux still > > gets the kbd IRQs. > > > > > just for testing. If we understand well, returning from the ISR with > > > RET_INTR_HANDLED should not propagate the > > > > > > IRQ to the other domains, therefore to the Linux one, right? > > > > > > Why can we then keep working with the shell? (I expect that we can not > > > interact with the shell, even not see the echoed char.). > > > > > > > > > > > > Does Xenomai patch the keyb driver? > > > > > > > No. > > > > > > > > > > > (We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > > > the /proc/interrupts, as expected). > > > > > > > > > > > > Thanks for your help > > > > > > > > > > > > Daniel > > > > > > > > > ------------------------------------------------------------------------ > > > > > > _______________________________________________ > > > Xenomai-help mailing list > > > Xenomai-help@domain.hid > > > https://mail.gna.org/listinfo/xenomai-help > > > > Thanks for your help Guillaume Boutillier > pièce jointe message de courriel, "Message transféré - Re: > [Xenomai-help] IRQ1 forwarded to Linux?" > Le jeudi 27 avril 2006 à 15:12 +0200, Sébastien Gerber a écrit : > > ROSSIER Daniel wrote: > > > Hello, > > > > > > > > > > > > We have a small Xeno application which implements an interrupt service > > > (ISR) on IRQ 1 (keyboard) – quite stupid, but > > > > > > > What does the following say? > > $ cat /proc/ipipe/Linux > > > > > just for testing. If we understand well, returning from the ISR with > > > RET_INTR_HANDLED should not propagate the > > > > > > IRQ to the other domains, therefore to the Linux one, right? > > > > > > Why can we then keep working with the shell? (I expect that we can not > > > interact with the shell, even not see the echoed char.). > > > > > > > > > > > > Does Xenomai patch the keyb driver? > > > > > > > > > > > > (We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > > > the /proc/interrupts, as expected). > > > > > > > > > > > > Thanks for your help > > > > > > > > > > > > Daniel > > > > > > > > > ------------------------------------------------------------------------ > > > > > > _______________________________________________ > > > Xenomai-help mailing list > > > Xenomai-help@domain.hid > > > https://mail.gna.org/listinfo/xenomai-help > > > > ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Xenomai-help] IRQ1 forwarded to Linux? 2006-04-27 14:44 ` Guillaume Boutillier @ 2006-04-27 14:56 ` Philippe Gerum 2006-04-27 15:36 ` Guillaume Boutillier 0 siblings, 1 reply; 6+ messages in thread From: Philippe Gerum @ 2006-04-27 14:56 UTC (permalink / raw) To: Guillaume Boutillier; +Cc: xenomai Guillaume Boutillier wrote: > Hello, > > >>>ROSSIER Daniel wrote: >>> >>>>Hello, >>>> >>>> >>>> >>>>We have a small Xeno application which implements an interrupt service >>>>(ISR) on IRQ 1 (keyboard) – quite stupid, but >>>> >>> >>>Sorry, >>> >>>- $ cat /proc/ipipe/Linux >>>+ $ cat /proc/ipipe/Xenomai > > > I checked when I run my application... > But I couldn't understand the meaning of the different flags : accepted, > passed and grabbed?! > What about posting the output to this list? Maybe someone could help. > >>>when your test is loaded would be useful to understand why Linux still >>>gets the kbd IRQs. >>> >>> >>>>just for testing. If we understand well, returning from the ISR with >>>>RET_INTR_HANDLED should not propagate the >>>> >>>>IRQ to the other domains, therefore to the Linux one, right? >>>> >>>>Why can we then keep working with the shell? (I expect that we can not >>>>interact with the shell, even not see the echoed char.). >>>> >>>> >>>> >>>>Does Xenomai patch the keyb driver? >>>> >>> >>>No. >>> >>> >>>> >>>> >>>>(We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on >>>>the /proc/interrupts, as expected). >>>> >>>> >>>> >>>>Thanks for your help >>>> >>>> >>>> >>>>Daniel >>>> >>>> >>>>------------------------------------------------------------------------ >>>> >>>>_______________________________________________ >>>>Xenomai-help mailing list >>>>Xenomai-help@domain.hid >>>>https://mail.gna.org/listinfo/xenomai-help >>> >>> > Thanks for your help > > Guillaume Boutillier > > >>pièce jointe message de courriel, "Message transféré - Re: >>[Xenomai-help] IRQ1 forwarded to Linux?" >>Le jeudi 27 avril 2006 à 15:12 +0200, Sébastien Gerber a écrit : >> >>>ROSSIER Daniel wrote: >>> >>>>Hello, >>>> >>>> >>>> >>>>We have a small Xeno application which implements an interrupt service >>>>(ISR) on IRQ 1 (keyboard) – quite stupid, but >>>> >>> >>>What does the following say? >>>$ cat /proc/ipipe/Linux >>> >>> >>>>just for testing. If we understand well, returning from the ISR with >>>>RET_INTR_HANDLED should not propagate the >>>> >>>>IRQ to the other domains, therefore to the Linux one, right? >>>> >>>>Why can we then keep working with the shell? (I expect that we can not >>>>interact with the shell, even not see the echoed char.). >>>> >>>> >>>> >>>>Does Xenomai patch the keyb driver? >>>> >>>> >>>> >>>>(We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on >>>>the /proc/interrupts, as expected). >>>> >>>> >>>> >>>>Thanks for your help >>>> >>>> >>>> >>>>Daniel >>>> >>>> >>>>------------------------------------------------------------------------ >>>> >>>>_______________________________________________ >>>>Xenomai-help mailing list >>>>Xenomai-help@domain.hid >>>>https://mail.gna.org/listinfo/xenomai-help >>> >>> > > > _______________________________________________ > Xenomai-help mailing list > Xenomai-help@domain.hid > https://mail.gna.org/listinfo/xenomai-help > -- Philippe. ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Xenomai-help] IRQ1 forwarded to Linux? 2006-04-27 14:56 ` Philippe Gerum @ 2006-04-27 15:36 ` Guillaume Boutillier 2006-04-27 16:20 ` Philippe Gerum 0 siblings, 1 reply; 6+ messages in thread From: Guillaume Boutillier @ 2006-04-27 15:36 UTC (permalink / raw) To: Philippe Gerum; +Cc: xenomai > Guillaume Boutillier wrote: > > Hello, > > > > > >>>ROSSIER Daniel wrote: > >>> > >>>>Hello, > >>>> > >>>> > >>>> > >>>>We have a small Xeno application which implements an interrupt service > >>>>(ISR) on IRQ 1 (keyboard) – quite stupid, but > >>>> > >>> > >>>Sorry, > >>> > >>>- $ cat /proc/ipipe/Linux > >>>+ $ cat /proc/ipipe/Xenomai > > > > > > I checked when I run my application... > > But I couldn't understand the meaning of the different flags : accepted, > > passed and grabbed?! > > > > What about posting the output to this list? Maybe someone could help. > Sorry I forgot the results... when I run the application with RT_INTR_PROPAGATE : $ cat /proc/ipipe/Xenomai Priority=200, Id=0x58454e4f irq0-1: grabbed irq2-31: passed irq32-33: passed, virtual irq34: grabbed, virtual $ cat /proc/ipipe/Linux Priority=100, Id=0x00000000 irq0-15: accepted irq32-33: grabbed, virtual irq34: passed, virtual when I run the application with RT_INTR_HANDLED $ cat /proc/ipipe/Xenomai Priority=200, Id=0x58454e4f irq0-1: grabbed irq2-31: passed irq32-33: passed, virtual irq34: grabbed, virtual $ cat /proc/ipipe/Linux Priority=100, Id=0x00000000 irq0-15: accepted irq32-33: grabbed, virtual irq34: passed, virtual Thanks Guillaume > > > >>>when your test is loaded would be useful to understand why Linux still > >>>gets the kbd IRQs. > >>> > >>> > >>>>just for testing. If we understand well, returning from the ISR with > >>>>RET_INTR_HANDLED should not propagate the > >>>> > >>>>IRQ to the other domains, therefore to the Linux one, right? > >>>> > >>>>Why can we then keep working with the shell? (I expect that we can not > >>>>interact with the shell, even not see the echoed char.). > >>>> > >>>> > >>>> > >>>>Does Xenomai patch the keyb driver? > >>>> > >>> > >>>No. > >>> > >>> > >>>> > >>>> > >>>>(We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > >>>>the /proc/interrupts, as expected). > >>>> > >>>> > >>>> > >>>>Thanks for your help > >>>> > >>>> > >>>> > >>>>Daniel > >>>> > >>>> > >>>>------------------------------------------------------------------------ > >>>> > >>>>_______________________________________________ > >>>>Xenomai-help mailing list > >>>>Xenomai-help@domain.hid > >>>>https://mail.gna.org/listinfo/xenomai-help > >>> > >>> > > Thanks for your help > > > > Guillaume Boutillier > > > > > >>pièce jointe message de courriel, "Message transféré - Re: > >>[Xenomai-help] IRQ1 forwarded to Linux?" > >>Le jeudi 27 avril 2006 à 15:12 +0200, Sébastien Gerber a écrit : > >> > >>>ROSSIER Daniel wrote: > >>> > >>>>Hello, > >>>> > >>>> > >>>> > >>>>We have a small Xeno application which implements an interrupt service > >>>>(ISR) on IRQ 1 (keyboard) – quite stupid, but > >>>> > >>> > >>>What does the following say? > >>>$ cat /proc/ipipe/Linux > >>> > >>> > >>>>just for testing. If we understand well, returning from the ISR with > >>>>RET_INTR_HANDLED should not propagate the > >>>> > >>>>IRQ to the other domains, therefore to the Linux one, right? > >>>> > >>>>Why can we then keep working with the shell? (I expect that we can not > >>>>interact with the shell, even not see the echoed char.). > >>>> > >>>> > >>>> > >>>>Does Xenomai patch the keyb driver? > >>>> > >>>> > >>>> > >>>>(We've seen that on /proc/xenomai/irq, we get the IRQ1 count, but NOT on > >>>>the /proc/interrupts, as expected). > >>>> > >>>> > >>>> > >>>>Thanks for your help > >>>> > >>>> > >>>> > >>>>Daniel > >>>> > >>>> > >>>>------------------------------------------------------------------------ > >>>> > >>>>_______________________________________________ > >>>>Xenomai-help mailing list > >>>>Xenomai-help@domain.hid > >>>>https://mail.gna.org/listinfo/xenomai-help > >>> > >>> > > > > > > _______________________________________________ > > Xenomai-help mailing list > > Xenomai-help@domain.hid > > https://mail.gna.org/listinfo/xenomai-help > > > > ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Xenomai-help] IRQ1 forwarded to Linux? 2006-04-27 15:36 ` Guillaume Boutillier @ 2006-04-27 16:20 ` Philippe Gerum 0 siblings, 0 replies; 6+ messages in thread From: Philippe Gerum @ 2006-04-27 16:20 UTC (permalink / raw) To: Guillaume Boutillier; +Cc: xenomai Guillaume Boutillier wrote: >>Guillaume Boutillier wrote: >> >>>Hello, >>> >>> >>> >>>>>ROSSIER Daniel wrote: >>>>> >>>>> >>>>>>Hello, >>>>>> >>>>>> >>>>>> >>>>>>We have a small Xeno application which implements an interrupt service >>>>>>(ISR) on IRQ 1 (keyboard) – quite stupid, but >>>>>> >>>>> >>>>>Sorry, >>>>> >>>>>- $ cat /proc/ipipe/Linux >>>>>+ $ cat /proc/ipipe/Xenomai >>> >>> >>>I checked when I run my application... >>>But I couldn't understand the meaning of the different flags : accepted, >>>passed and grabbed?! >>> >> >>What about posting the output to this list? Maybe someone could help. >> > > > Sorry I forgot the results... > > when I run the application with RT_INTR_PROPAGATE : > > $ cat /proc/ipipe/Xenomai > Priority=200, Id=0x58454e4f > irq0-1: grabbed > irq2-31: passed > irq32-33: passed, virtual > irq34: grabbed, virtual > > $ cat /proc/ipipe/Linux > Priority=100, Id=0x00000000 > irq0-15: accepted > irq32-33: grabbed, virtual > irq34: passed, virtual > > > when I run the application with RT_INTR_HANDLED > > $ cat /proc/ipipe/Xenomai > Priority=200, Id=0x58454e4f > irq0-1: grabbed > irq2-31: passed > irq32-33: passed, virtual > irq34: grabbed, virtual > > $ cat /proc/ipipe/Linux > Priority=100, Id=0x00000000 > irq0-15: accepted > irq32-33: grabbed, virtual > irq34: passed, virtual > "grabbed" means that the domain intercepts the IRQ, and will let its own handler decide whether the interrupt needs to be propagated down the pipeline or not (a manual call to ipipe_propagate_irq will be needed if so). IOW, Adeos stops scanning the pipeline chain upon IRQ receipt after this domain's stage; it would do that otherwise to discover new (lower-priority) domains wanting to handle this IRQ too. "accepted" means that Adeos will let the domain process the IRQ, and keep on searching for (lower-priority) domains wanting to handle it too. "passed" means that Adeos will not ask the domain to handle the IRQ, but will try to pass it to the next domain down the pipeline. "discarded" means to make as if the IRQ never happened starting from the current one, i.e. neither the current nor the lower-priority domain will ever know about it. Regarding the specific case of the keyboard input still working despite the kbd interrupts can't flow, it's likely due to the fact that the input sub-system (i8042 serio driver) also polls the attached devices according to a HZ/20 period, in addition to handling the interrupts. -- Philippe. ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2006-04-27 16:20 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-04-27 11:17 [Xenomai-help] IRQ1 forwarded to Linux? ROSSIER Daniel
2006-04-27 11:57 ` Philippe Gerum
2006-04-27 11:59 ` Philippe Gerum
[not found] <1146143577.3671.3.camel@domain.hid>
2006-04-27 14:44 ` Guillaume Boutillier
2006-04-27 14:56 ` Philippe Gerum
2006-04-27 15:36 ` Guillaume Boutillier
2006-04-27 16:20 ` Philippe Gerum
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.