On 3/29/07, Erik Mouw <mouw@xxxxxxxxxxxx> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thu, Mar 29, 2007 at 12:57:39PM +0530, sandeep lahane wrote: > On 3/29/07, Rajat Jain <rajat.noida.india@xxxxxxxxx> wrote: [trim lots of irrelevant quoted text] > >2) When the IRQ occurs, the kernel invokes EACH and every ISR hooked > >onto that IRQ (Not just the ISR that actually services the device). > I think this is one of the reasons why Linux is not a real time OS, > please CMIIW. No, that has nothing to do with it because this is a hardware limitation. The only way to figure out which device triggered the IRQ is to check each individual device connected to that IRQ line. Because checking the interrupt status is highly device specific, the only sensible way would be to let the ISR for the device figure out. You always have to do that, even with an RTOS. Erik - -- They're all fools. Don't worry. Darwin may be slow, but he'll eventually get them. -- Matthew Lammers in alt.sysadmin.recovery -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFGC6+u/PlVHJtIto0RAgj8AJ0epGG+hSLC1cTt2/eNEockiWJI+gCfdS0U Hv+LcWE3RVidHSXeopg1EvA= =FpaM -----END PGP SIGNATURE-----
That means this is not the one of the sources of causing more interrupt latency, and this can not be removed? RTOSes also do the same thing since this is h/w specific. Cool, makes lot of sense, need to explore it more. -- Regards, Sandeep. -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
