Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Matthew Dillon
:> The assumption that an I/O port address is lower than 0x8000 is OK for :> IBM-PC's ISA-bus, but PC98 can use hiher addresses for ISA-bus like :> bus. : :I understand. I think the best thing is to store the port address as an :int. I don't think there is a need to fork the file into a pc98 versio

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Mikhail Teterin
Doug Rabson once stated: => If this should not be in sys/i386/isa, I will copy this file into => sys/pc98/pc98. Comment please. = =It needs to be signed since the value -1 is used to indicate no port =resources. Perhaps it would be better to change it to 'int' instead of ='u_short'. Uff, waistin

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Mike Smith
> > Singed shorts are often the result of sitting on a stove during the > summer. > > We also recently saw a "hanging root device to..." in a boot on one of our > machines. You'll see the 'c' up above all of the SCSI device probe messages. Not sure how we go about making that one tidy. --

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Doug Rabson
On Sun, 25 Apr 1999, KATO Takenori wrote: > Doug Rabson wrote: > > > It needs to be signed since the value -1 is used to indicate no port > > resources. Perhaps it would be better to change it to 'int' instead of > > 'u_short'. > > Oops, I didn't realize this case. > > The assumption that an I

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Robert Watson
Singed shorts are often the result of sitting on a stove during the summer. We also recently saw a "hanging root device to..." in a boot on one of our machines. :-) Robert N Watson rob...@fledge.watson.org http://www.watson.org/~robert/ PGP key fingerprint: AF B5 5F FF A6 4A

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread KATO Takenori
Doug Rabson wrote: > It needs to be signed since the value -1 is used to indicate no port > resources. Perhaps it would be better to change it to 'int' instead of > 'u_short'. Oops, I didn't realize this case. The assumption that an I/O port address is lower than 0x8000 is OK for IBM-PC's ISA-b

Re: singed short to unsigned long conversion problem in isa.c

1999-04-24 Thread Doug Rabson
On Sat, 24 Apr 1999, KATO Takenori wrote: > The type of the member id_port in isa_device structure is signed short > and a value is converted int unsigned long to allocate I/O port > resource via bus_alloc_resources. If an I/O port >= 0x8000, > conversion from signed short to unsigned long cause