[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ARMv4-support in armel/squeeze?

> FWIW, my box is a "Raidsonic NAS IB-4220-B" (machtype 2038, FA526 CPU,

IIRC these are not strongarm based. They're a cheap NAS chipset from a 
Tiwanese manufacturer who is still knocking out custom cores that only 
implement ARMv4. I assume that being prehistoric and effectively incompatible 
with every other ARM core in production use, it's cheaper to licence.

> > It's very difficult to support people with reliable old Netwinders at
> > one end, and others who want to use all the shiny stuff on their new
> > arm netbooks/tablets at the other witout building 6 different arm
> > distros. What an appropriate compromise is, is currently a hot topic.
> Um, really six (the ARM world is new to me so I might miss irony)?

Six is a fairly conservative estimate. I count at least ten significant 
variants off the top of my head.  v4, v4t, v5te, v6, v7. Plus variants using 
Thumb and 2-4 different flavors of VFP/NEON.  And that's before we start 
doubling VFP for both soft and hard float ABIs.

> Just to improve my understanding, how bad was the impact of a
> hypothetical "armv4" architecture on ARMv4t, i.e. without thumb
> support?

Thumb support is not optional under the EABI[1].  You could make a separate 
incompatible port for armv4 only, but realistically that's about the same 
amount of work and probably even less users than the old "arm" port.

If you want to support both EABI and armv4 with in the same binary then this 
incurs a runtime cost of at least one (often two) additional branch and 
compare instructions per function invocation.


[1] This isn't just a nominal compliance issue.  Lack of Thumb support 
severely limits your options for later (v5+, especially v7) hardware.

Reply to: