| Re: [eigen] Eigen in ARM Cortex M3? |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [eigen] Eigen in ARM Cortex M3?
- From: Benoit Jacob <jacob.benoit.1@xxxxxxxxx>
- Date: Mon, 14 Nov 2011 07:54:53 -0500
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; bh=lVK86W1zEVORfkcHLBksJfCpzQwurB4oQbfb8KYxWcI=; b=XXZlGrMSEZ/JFwriAUifeDVYtZUVShNYzugGy+hfhOS5tUopPQe4vBIxSxlAGqHSIN 7Xd+nEDaF6THvPLPzOG2iW9NgEcgRUTGInUhZG9BREP+q/7zPpt2lOXY/mQ5vR4GdQ4l QQ58Tsm0HOpTiG62KG6LAHkqQ4LVtmS7bVr2I=
2011/11/14 Gael Guennebaud <gael.guennebaud@xxxxxxxxx>:
>> without an OS, I'd say no mainly because we need a memory allocator.
>
> well, if you use fixed sized objects only then you can still provide a
> dummy malloc() that always return 0 just to avoid linking issues.
Even if they have dynamic-sized matrices, they can still have a memory
allocator implementing malloc(), without an OS...
Think of it this way: there already as malloc() in Borland Turbo C for
MS-DOS ;-)
No OS just means that the memory allocator has to deal directly with
allocating memory at the lowest level, instead of just mmapping pages.
Benoit