9:29 a.m.
Derek,
It is less "powers that be" than a collaborative effort/decision. We do not have
resources to justify keeping v5 alive but you are free to coordinate with others and pick
it up, in the same way that Seneca are to own v6 support (maybe Seneca can even help with
build system setup if you ask them). Do you have any interest in driving that?
You will find the ominous powers that be are in fact a bunch of us doing the work who are
overloaded enough to keep just v7 and v8 on track :) For those who are devastated and have
no v7 hardware, ping me off list and maybe I can look into getting a couple of v7 boards
out there.
Jon.
--
Sent from my phone. Please excuse brevity.
Derek Atkins <warlord(a)MIT.EDU> wrote:
Quentin Armitage <quentin(a)armitage.org.uk> writes:
since there has been no major objection i will disable building
armv5tel rpms in rawhide before the mass rebuild.
Dennis
I guess it's too late now, but I got a few days behind on my list emails. I
use 2 * Sheevaplugs and 2 * Dreamplugs with Fedora, and would be very
disappointed to see support for them being dropped from Fedora. For me, I
still see quite a lifetime in them for what they are doing.
I've mentioned multiple times my hope to keep kirkwood support in
Fedora, but alas it feels like the powers that be just don't care about
us *plug users. :( If I want to continue using my plugs I guess I'll
have to learn Debuntu. :(
Quentin Armitage
-derek
--
Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
Member, MIT Student Information Processing Board (SIPB)
URL: http://web.mit.edu/warlord/ PP-ASEL-IA N1NWH
warlord(a)MIT.EDU PGP key available
_______________________________________________
arm mailing list
arm(a)lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/arm
10:20 a.m.
New subject: arm software floating point support going forward
Jon,
Thanks for the reply. I certainly don't have the hardware to host a v5
koji. All I own are 2 Sheevas and 2 Gurus (and not even the "Server
Plus" models). Moreover, one of my Sheevas is in "critical" production
on my network so I can't really pull it away to perform other duties.
I was actually looking at a Cubieboard to get some newer, more powerful
hardware, because the Guru is way too low powered to run my MySQL
instance. I may even find that the Cubie is too low-powered, but
obviously cannot test that until I have one or have access to one.
Regardless, right now I don't have the budget to acquire new hardware,
which is why I'd like to continue using what I already own.
Not being intimately familiar with the various changes in the hardware I
guess I just don't understand why we need so many target-specific
distributions? I thought the only issue was the floating point ABI
issue, which would lead me to believe that we only would need two, FP
and non-FP? Is there really a significant speed improvement with
e.g. v7 or v8 when compiled specifically vs. running e.g. v6 on a v7 or
v8? ISTR that measurements showed some but relatively insignificant
speed differences, so why not just stay at the lowest level to support
more hardware?
Thanks,
-derek
Jon Masters <jcm(a)redhat.com> writes:
Derek,
It is less "powers that be" than a collaborative effort/decision. We do not
have resources to justify keeping v5 alive but you are free to coordinate with others and
pick it up, in the same way that Seneca are to own v6 support (maybe Seneca can even help
with build system setup if you ask them). Do you have any interest in driving that?
You will find the ominous powers that be are in fact a bunch of us doing the work who are
overloaded enough to keep just v7 and v8 on track :) For those who are devastated and have
no v7 hardware, ping me off list and maybe I can look into getting a couple of v7 boards
out there.
Jon.
--
Sent from my phone. Please excuse brevity.
Derek Atkins <warlord(a)MIT.EDU> wrote:
Quentin Armitage <quentin(a)armitage.org.uk> writes:
> since there has been no major objection i will disable building
> armv5tel rpms in rawhide before the mass rebuild.
>
> Dennis
>
> I guess it's too late now, but I got a few days behind on my list emails. I
> use 2 * Sheevaplugs and 2 * Dreamplugs with Fedora, and would be very
> disappointed to see support for them being dropped from Fedora. For me, I
> still see quite a lifetime in them for what they are doing.
I've mentioned multiple times my hope to keep kirkwood support in
Fedora, but alas it feels like the powers that be just don't care about
us *plug users. :( If I want to continue using my plugs I guess I'll
have to learn Debuntu. :(
> Quentin Armitage
-derek
--
Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
Member, MIT Student Information Processing Board (SIPB)
URL: http://web.mit.edu/warlord/ PP-ASEL-IA N1NWH
warlord(a)MIT.EDU PGP key available
8:52 a.m.
New subject: arm software floating point support going forward
On Fri, Feb 1, 2013 at 4:20 PM, Derek Atkins <warlord(a)mit.edu> wrote:
Jon,
Thanks for the reply. I certainly don't have the hardware to host a v5
koji. All I own are 2 Sheevas and 2 Gurus (and not even the "Server
Plus" models). Moreover, one of my Sheevas is in "critical" production
on my network so I can't really pull it away to perform other duties.
I was actually looking at a Cubieboard to get some newer, more powerful
hardware, because the Guru is way too low powered to run my MySQL
instance. I may even find that the Cubie is too low-powered, but
obviously cannot test that until I have one or have access to one.
Regardless, right now I don't have the budget to acquire new hardware,
which is why I'd like to continue using what I already own.
Not being intimately familiar with the various changes in the hardware I
guess I just don't understand why we need so many target-specific
distributions? I thought the only issue was the floating point ABI
issue, which would lead me to believe that we only would need two, FP
and non-FP? Is there really a significant speed improvement with
e.g. v7 or v8 when compiled specifically vs. running e.g. v6 on a v7 or
v8? ISTR that measurements showed some but relatively insignificant
speed differences, so why not just stay at the lowest level to support
more hardware?
Basically ARMv7 is the current generation of chipsets. ARMv5 and ARMv6
is basically dead. ARMv5 SoCs are end of line in ARM and as they
require things like DDR1 they are getting expensive for manufacturers
to produce so available hardware is dwindling. They also only support
a max of 1gb of ram and are limited. ARM is moving all the remaining
users of v5 onto platforms like Cortex-A5 chips that use more modern
memory and are cheaper to produce.
There have never been that many ARMv5 platforms suitable to running
fedora with the kirkwood plugs basically being it, the HW isn't that
capable and availability will only shink with time. It's not an
insignificant amount of effort to support this platform. I think I as
the person that basically does all the kernel maintenance and all the
building of packages and the vast majority of fixing of the
problematic packages or chasing people to fix them I think I'm one of
the few that is able to quantify this time and effort.
To go with that there's frankly not a lot of people actually bothering
to test armv5 and this is categorised with the fact that some core
packages and functionality were broken in F-18 for quite some time and
not even reported so issues like that helped us make up our mind
whether it was critical to test.
F-17 download numbers I don't believe were large.
Ultimately we need to assess where our efforts are best directed to
get our best bang for our buck. Most of those that are doing
significant work on the platform are volunteers. The fact is that
while it is sad for us to be dropping the support I don't believe it's
the best way for those people that are doing their work for them to
invest their time for a platform that ARM themselves is dead with
limited machines available now and their availability in new products
moving forward it's basically extinct.
Regards,
Peter
4126
days inactive
4127
days old
2 comments
3 participants
participants (3)
-
Derek Atkins -
Jon Masters -
Peter Robinson