Re: [hatari-devel] symbols not always loaded when entering debugger ? |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/hatari-devel Archives
]
- To: hatari-devel@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [hatari-devel] symbols not always loaded when entering debugger ?
- From: David Savinkoff <dsavnkff@xxxxxxxxx>
- Date: Sat, 18 Nov 2017 14:15:57 -0700 (MST)
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=telus.net; s=neo; t=1511039758; bh=f5cA8KAHDXl6z22oKmUjlwWMCqCh1aQy4qHBFe5Kz6k=; h=Date:From:To:In-Reply-To:Subject; b=ln8xFSEZ81/7JTdK1jufHGr1gMIkBdfNewXnGCoyGCypBWQxmGN7gKbgsbKSkwL8E rNfjosnXuB+n6j/l6ZvYepVKKkj/TOxQikbkBMsMiubBY82CLT/FD1RWCdL4RKbWJy CSMSwF5jjC1JlrtBiM4v21V7fLaSGzR8KLtmdPbvNehuzsnghW9+klmMjO8cCeSw6s j1uAsxs3i+TvIhF3rDNU5yWwvzXahwT77hKTXZRJ0oB78nBfLpO1SqkNaUcCqXGh1e PWuW4kAlqEc8ZOR9Nbv6e8tW3nT7icZejV8nXj+3XeVA1qh1YVIhuEeNAir4lukV8o fI3yTMf7gnBpw==
- Thread-index: U6KVM3q6zvz2bkys91ame4yptMRA/g==
- Thread-topic: symbols not always loaded when entering debugger ?
Hatari has so many options that higher level options are needed.
These higher level options should be default 'user friendly' RTFM_not
(see ffmpeg as an example). I don't know what a call-graph is, and
that part of the manual scares me.
----- On Nov 17, 2017, at 3:47 PM, Eero Tamminen wrote:
> Hi,
>
> On 11/18/2017 12:34 AM, Nicolas Pomarède wrote:
>> Le 17/11/2017 à 22:43, Eero Tamminen a écrit :
>> Yes, but my problem is "simple" : I want to code a program for STF/STE
>> that will run on real HW. I don't have specific memory or OS needs, but
>> it's just that in the past there where some difference between TOS and
>> emutos about HW init (for example microwire init could be different).
>
> I confess, I had in mind also potential improvements to EmuTOS. :-)
>
>
>> I prefer to work the other way around : make sure the program works with
>> real TOS and then in the end check it works with emutos too.
>
> Ok.
>
>
> [...]
>>> I think I'll provide an extra script that invokes Hatari
>>> with options that configure it more suitably for debugging.
>>> It can be installed alongside Hatari.
>>>
>>> Would "hatari-debug" be OK name for it?
>>>
>>> Besides the "symbols resident" thing, I think by default it
>>> should use "--trace os_base", and have some shorter option
>>> (-e?) for "--debug-except".
>>>
>>> It can output some debugger usage info (e.g. note about
>>> "setopt -D"), and start a terminal for Hatari output if
>>> user didn't start it from terminal.
>>>
>>> Anything else?
>>
>> But why yet another script and some shorter options when it would be
>> much simpler to make the above default in normal Hatari ?
>
> Before it's going to be default, it needs to be tested
> (couple of weeks of use with Atari binaries linked with
> different libs & linkers should be enough).
>
> Such a script will help you and other debugger users in
> doing that.
>
> However, I think something like that could be also of more
> general use. So, please consider what such a script
> could & should do.
>
>
>> If you agree that another script could be needed to make debugging
>> easier in the scenario I reported, then the simplest would be to change
>> the default to keep symbols resident.
>
> By extending the same reasoning, why Hatari shouldn't
> invoke debugger on every exception?
>
> (Debugging use-cases are different from non-debugging
> use-cases and debugging settings can have downsides for
> normal emulator usage. While not as drastic, automatic
> symbol loading for terminating programs can also have
> downsides, at least it generates extra output to console.)
>
>
>> I'm afraid people won't read doc or look for what such or
>> such script is doing.
>
> I wonder how they happened to run Hatari in the first place.
> By accidentally installing and running Hatari, when their
> cat ran over the keyboard, perhaps? :-)
>
>
> - Eero