Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 03 Dec 2007 21:27:40 -0800
From:      Julian Elischer <julian@elischer.org>
To:        Marcel Moolenaar <xcllnt@mac.com>
Cc:        FreeBSD Current <current@freebsd.org>
Subject:   Re: remote gdb failures in recent builds (last 6 months or so.)
Message-ID:  <4754E54C.20104@elischer.org>
In-Reply-To: <A3A1F0CB-CF92-4D42-8211-241A561D2130@mac.com>
References:  <4754D411.6060608@elischer.org> <A3A1F0CB-CF92-4D42-8211-241A561D2130@mac.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Marcel Moolenaar wrote:
> 
> On Dec 3, 2007, at 8:14 PM, Julian Elischer wrote:
> 
>>
>> Here's a trace of kgdb attempting to sync with a kernel over the 
>> serial port..
>>
>> note that:
>> 1/ gdb CAN read the damned data.. this is obtained from ktrace.
>> 2/ it prints out the packet in the error message.
>> 3/ teh kernel is sending each message twice, in pairs.
>> 4/ after kgdb finally sends a '-' (NAK) teh kernel sends ONE MORE
>> which seems to be accepted
>>
>> Anyone have any clues?
> 
> Works just fine for me on all platforms, but I always use uart(4).
> 
> What platform are you using and what gdb?
> How's the serial connection; through a comm-server?
> What's the line setting, which driver?

sio driver on Dell 2650 running new x86 kernel.. 
standard com1-com2 type ports.

Both systems as installed from source. (-current plus or minus a month)
null-modem seraial cable between the ports.
port on target is sio1. flags are 0xc0 (reserve for gdb and use for gdb).

trace in previous mail is extracted from gdb via ktrace, so it is getting 
all the bytes but somehow failing to process them.

It's slowly been getting worse (across many pairs of machines) over 
the last few years. finally today it just doesn't work at all. I've used 
this pair of machines many times before. usually it takes "luck" to get 
synced up but once it's synced, it's solid.



> 




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4754E54C.20104>