Galileo GT-64260: SERR* and machine check
From: sk (ivdal_at_yahoo.com)
Date: 12/23/04
- Next message: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Previous message: Alan Kilian: "Re: Problem in using compiling C++ project"
- Next in thread: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Reply: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: 23 Dec 2004 09:23:03 -0800
Hi,
Look like I have a problem here..
When the SERR* line on the PCI bus is asserted, the application
which was performing the transaction is supposed to get a SIGBUS,
right? What is the right way for the OS to find out that SERR* was
asserted besides polling the status register in the host bridge?
I expected the host bridge to generate a machine check exception
or something (on PowerPC, that is), but it looks like GT-64260A
just ignores the SERR* - no exception, no machine check. I scoured
the docs but found no relevant hints.
The host bridge is GT-64260A. The board is MVME5500.
The CPU is PowerPC 7457.
Thanks, all
DM
- Next message: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Previous message: Alan Kilian: "Re: Problem in using compiling C++ project"
- Next in thread: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Reply: Mikael Pettersson: "Re: Galileo GT-64260: SERR* and machine check"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]