[Apologies for the wide distibution of this mail (Debian, Ubuntu and Fedora main+arm dev lists, plus linaro and lsb) but it's useful to catch people who care about this issue enough to do some work. Do please bear the distribution in mind when replying, focussing any detailed discussion on linaro-dev please. I'll post a summary at the end if there is material input]
---------------
The subject of the Linux Standards Base (LSB) and ARM came up at the recent Linaro Dev summit.
During discussion of standardisation of ABI across distributions (Ubuntu, Debian, Fedora etc) it was suggested that maybe the LSB was a useful place to specify some kind of agreed minimum.
It turns out that the LSB supports 7 architectures, but does not include ARM, beyond the catch-all 'generic'. This seemed an odd omission so I contacted the LSB people who were very helpful, and I found that they would like to support ARM but a) there was not a clear binary ABI standard to support in the past (there were lots of variants) and b) no-one really stepped up to do the work of porting the LSB docs and tools.
It came up on the LSB list too: https://lists.linux-foundation.org/pipermail/lsb-discuss/2011-May/006828.htm...
As I say in that thread, I think a) has been dealt with in that there is now an agreed base with wide-enough support that we could usefully specify (the armv7, VFP-D16 ('hard-float'), little-endian ABI, as used by Debian 'armhf', Ubuntu 'armhf', Fedora 'armv7hl', and also Meego 'armv7hl'). That fits with what Linaro is supporting too.
Mats D Wichman kindly gave some idea of how much work in needed to 'port' the LSB in the above thread:
------------- in that range of options [week, month, year, 6 years], it's closer to a year than to any of the others.
The spec work isn't hard by itself if there's a reasonable processor-level ABI document available, which unless something has changed, is the case for ARM: like the psABI documents that exist for other architectures, reference is made to such a base document (or set) for things like register assignments, calling conventions, exception processing, etc. so it doesn't have to be written - maybe just pinning down where the base document offers choices, that the LSB ABI does it this way.
In addition to the spec, there's a great deal of code around LSB, which all has to be adapted. It's obviously portable code since it works for seven architectures already, of varying wordsizes, endian-ness, etc. Where there are details specific to a processor architecture, we actually have all of the details stored in a database (which can be browsed at http://dev.linuxfoundation.org/navigator), and the biggest task actually becomes populating the database with data for this new architecture. There are some fairly reasonable tools for scanning a distribution which would provide a useful starting point, but then someone has to validate that things are all correct. Some of the validation happens by building and running iteratively various checkers which are part of the software suite anyway. That will require adjusting a number of makefiles, populating new trees under arch-specific names, etc. but that part is easy enough, just manual work. It has been rather a long time since a new architecture was added, I think PPC64 and S390X were added at about the same time and it was many years ago), so the procedure hasn't really been tested out recently.
Then there are a bunch of test suites which need to run to validate that a distribution conforms, and in my experience, this ends to be where new issues show up that break the assumption that everything's clean and portable, so there may well be extra debugging here. ------------------
As this is a non-trivial amount of work, the question then arises, does anyone care about this enough to actually do the work? Linaro is an obvious organisation that could expend some engineering effort on this, but to do that it needs some indication that it's more than a 'would-be-nice'.
Who actually uses the LSB for making widely-distributed binaries? Would anyone do so on ARM if it was specced? Is it important to make ARM a 'real' architecture alongside the others, e.g. especially in server space?
In my experience anyone distributing binaries actually picks a small set of distros and builds for those explicitly, rather than relying on the LSB. Does that mean that it's not actually useful in the real world? I guess in a sense this posting is to the wrong lists; we're all free software people here who have little use for the LSB. Where do the proprietary software distributors hang out :-)
It's easy to think of potential use-cases, and I think ultimately, unless the LSB is in fact entirely irrelevant, this work will get done everntually. But should we get on with that now, rather than whatever else we might be fixing, and if so, who is volunteering to get involved?
( Jon Masters and I have both expressed interest but are not exactly brimming over with spare time. Any more for any more?)
Opinions welcome.
Wookey
Wookey, the short answer is 'yes'. The next question is 'who?'. Maybe this can be bolted onto the hard float work, I'll let Konstantinos and Steve respond... Dave
On 06/01/11 01:22, Wookey wrote:
[Apologies for the wide distibution of this mail (Debian, Ubuntu and Fedora main+arm dev lists, plus linaro and lsb) but it's useful to catch people who care about this issue enough to do some work. Do please bear the distribution in mind when replying, focussing any detailed discussion on linaro-dev please. I'll post a summary at the end if there is material input]
The subject of the Linux Standards Base (LSB) and ARM came up at the recent Linaro Dev summit.
During discussion of standardisation of ABI across distributions (Ubuntu, Debian, Fedora etc) it was suggested that maybe the LSB was a useful place to specify some kind of agreed minimum.
It turns out that the LSB supports 7 architectures, but does not include ARM, beyond the catch-all 'generic'. This seemed an odd omission so I contacted the LSB people who were very helpful, and I found that they would like to support ARM but a) there was not a clear binary ABI standard to support in the past (there were lots of variants) and b) no-one really stepped up to do the work of porting the LSB docs and tools.
It came up on the LSB list too: https://lists.linux-foundation.org/pipermail/lsb-discuss/2011-May/006828.htm...
As I say in that thread, I think a) has been dealt with in that there is now an agreed base with wide-enough support that we could usefully specify (the armv7, VFP-D16 ('hard-float'), little-endian ABI, as used by Debian 'armhf', Ubuntu 'armhf', Fedora 'armv7hl', and also Meego 'armv7hl'). That fits with what Linaro is supporting too.
Mats D Wichman kindly gave some idea of how much work in needed to 'port' the LSB in the above thread:
in that range of options [week, month, year, 6 years], it's closer to a year than to any of the others.
The spec work isn't hard by itself if there's a reasonable processor-level ABI document available, which unless something has changed, is the case for ARM: like the psABI documents that exist for other architectures, reference is made to such a base document (or set) for things like register assignments, calling conventions, exception processing, etc. so it doesn't have to be written - maybe just pinning down where the base document offers choices, that the LSB ABI does it this way.
In addition to the spec, there's a great deal of code around LSB, which all has to be adapted. It's obviously portable code since it works for seven architectures already, of varying wordsizes, endian-ness, etc. Where there are details specific to a processor architecture, we actually have all of the details stored in a database (which can be browsed at http://dev.linuxfoundation.org/navigator), and the biggest task actually becomes populating the database with data for this new architecture. There are some fairly reasonable tools for scanning a distribution which would provide a useful starting point, but then someone has to validate that things are all correct. Some of the validation happens by building and running iteratively various checkers which are part of the software suite anyway. That will require adjusting a number of makefiles, populating new trees under arch-specific names, etc. but that part is easy enough, just manual work. It has been rather a long time since a new architecture was added, I think PPC64 and S390X were added at about the same time and it was many years ago), so the procedure hasn't really been tested out recently.
Then there are a bunch of test suites which need to run to validate that a distribution conforms, and in my experience, this ends to be where new issues show up that break the assumption that everything's clean and portable, so there may well be extra debugging here.
As this is a non-trivial amount of work, the question then arises, does anyone care about this enough to actually do the work? Linaro is an obvious organisation that could expend some engineering effort on this, but to do that it needs some indication that it's more than a 'would-be-nice'.
Who actually uses the LSB for making widely-distributed binaries? Would anyone do so on ARM if it was specced? Is it important to make ARM a 'real' architecture alongside the others, e.g. especially in server space?
In my experience anyone distributing binaries actually picks a small set of distros and builds for those explicitly, rather than relying on the LSB. Does that mean that it's not actually useful in the real world? I guess in a sense this posting is to the wrong lists; we're all free software people here who have little use for the LSB. Where do the proprietary software distributors hang out :-)
It's easy to think of potential use-cases, and I think ultimately, unless the LSB is in fact entirely irrelevant, this work will get done everntually. But should we get on with that now, rather than whatever else we might be fixing, and if so, who is volunteering to get involved?
( Jon Masters and I have both expressed interest but are not exactly brimming over with spare time. Any more for any more?)
Opinions welcome.
Wookey
On Wed, Jun 01, 2011 at 07:32:18AM +0900, David Rusling wrote:
On 06/01/11 01:22, Wookey wrote:
As this is a non-trivial amount of work, the question then arises, does anyone care about this enough to actually do the work? Linaro is an obvious organisation that could expend some engineering effort on this, but to do that it needs some indication that it's more than a 'would-be-nice'.
Wookey, the short answer is 'yes'. The next question is 'who?'.
Absolutely, yes. It makes a lot of sense for the growing number of people looking to collaborate here to pool their efforts in an existing central spec/location.
Maybe this can be bolted onto the hard float work, I'll let Konstantinos and Steve respond...
Ah, I guess I've just volunteered myself haven't I? :-)
/me goes to subscribe to the LSB lists and start reading things.
Cheers,
Steve, thanks, have a poke around and tell me / OCTO what it means to join in, what the costs and benefits are etc...
Dave
On 06/01/11 22:36, Steve McIntyre wrote:
On Wed, Jun 01, 2011 at 07:32:18AM +0900, David Rusling wrote:
On 06/01/11 01:22, Wookey wrote:
As this is a non-trivial amount of work, the question then arises, does anyone care about this enough to actually do the work? Linaro is an obvious organisation that could expend some engineering effort on this, but to do that it needs some indication that it's more than a 'would-be-nice'.
Wookey, the short answer is 'yes'. The next question is 'who?'.
Absolutely, yes. It makes a lot of sense for the growing number of people looking to collaborate here to pool their efforts in an existing central spec/location.
Maybe this can be bolted onto the hard float work, I'll let Konstantinos and Steve respond...
Ah, I guess I've just volunteered myself haven't I? :-)
/me goes to subscribe to the LSB lists and start reading things.
Cheers,
On Tue, May 31, 2011 at 5:22 PM, Wookey wookey@wookware.org wrote:
In my experience anyone distributing binaries actually picks a small set of distros and builds for those explicitly, rather than relying on the LSB. Does that mean that it's not actually useful in the real world? I guess in a sense this posting is to the wrong lists; we're all free software people here who have little use for the LSB. Where do the proprietary software distributors hang out :-)
the proprietary software distributors hang out around USA lawyer offices, where they get advice on how to perform tivoisation without anybody noticing. they then ship TVs and even 3G modems with embedded linux kernels and custom OSes... and nobody notices.
my take on this is that ARM is still just emerging from the "uselessness" of sub-600mhz ARM9s and ARM11s as far as general-purpose computing is concerned [laptop / desktop etc. *not* true embedded purposes obviously: don't get upset, ARM employees, because "mr LKCL said your processors were quotes useless quotes" - read it again: it's a *conditional* description]. also, the sheer diversity of SoCs plays directly, psychologically, against anyone "joining forces" on things like LSB. thus the majority of proprietary software distributors up until recently have been doing custom-built from scratch software stacks [using e.g. buildroot, openembedded] and thus LSB was and still is completely useless to them.
even android is custom-built, and everything (except the highly-optimised apps - for ARM - which are becoming more common) is a java app.
that having been said, 500mhz+ Dual-Core Cortex A9s already out which knock the stuffing out of 1.6ghz Intel Atoms (yes, saw the youtube video) mean that could just be about to change, completely.
sooo... although the situation *right now* is that nobody in the commercial world is the slightest bit interested in LSB because they all do "custom builds" of complete software stacks, it could be said that *if* the free software community just dropped ready-to-go LSB standards in front of their noses, they'd quite likely use it.
you have to remember that the majority of these companies could not put two lines of code together to save their lives. they literally have to be spoon-fed (in some cases even to the point of being told where to put the screws, let alone the software). they are usually spoon-fed by the CPU manufacturer [and in the case of MStar Semi, they won't even let *you* violate the GPL, they do it entirely for you].
so in that regard, i think it's more a case of "if the free software community provides LSB across ARM, it'll get used".
so in _that_ regard, the question becomes: "are the efforts of the free software community better off being spent elsewhere"? and "what benefit is there *TO THE FREE SOFTWARE COMMUNITY* of doing LSB for ARM"? forget the proprietary junkies, they'll suck anything from us that moves and not give a dime in return.
l.
On 06/01/2011 07:25 AM, Luke Kenneth Casson Leighton wrote:
so in _that_ regard, the question becomes: "are the efforts of the free software community better off being spent elsewhere"? and "what benefit is there *TO THE FREE SOFTWARE COMMUNITY* of doing LSB for ARM"? forget the proprietary junkies, they'll suck anything from us that moves and not give a dime in return.
That seems to be my cue to provide the community case for the LSB.
The LSB provides several things to the community:
- a framework for allowing Linux distributions to pool their userbase and work together as one platform instead of multiple platforms, one per distro
- test suites which identifies both compatibility problems and outright bugs to be detected and fixed
- a method for targeting builds at multiple distributions at once, both proprietary and free
- reporting tools for finding portability problems in built apps, again for both proprietary and free apps
We currently provide all of this for 7 architectures. ARM benefits indirectly (for example, many of the compatibility breaks detected on, say, x86_64 will affect all archs equally), but indirect support doesn't include the tools we've developed, and often compatibility issues are arch-specific.
At the risk of overstating the obvious, there are also ABI guarantees at stake here, which in my mind are architecture agnostic. OpenGL applications need to know which bits (API functions) of which core versions can be expected to be resolved during load time and which must be queried through GetProcAddress. LSB specifies this and makes it unambiguous.
cheers, Jesse
On Wed, Jun 1, 2011 at 8:00 AM, Jeff Licquia jeff@licquia.org wrote:
On 06/01/2011 07:25 AM, Luke Kenneth Casson Leighton wrote:
so in _that_ regard, the question becomes: "are the efforts of the free software community better off being spent elsewhere"? and "what benefit is there *TO THE FREE SOFTWARE COMMUNITY* of doing LSB for ARM"? forget the proprietary junkies, they'll suck anything from us that moves and not give a dime in return.
That seems to be my cue to provide the community case for the LSB.
The LSB provides several things to the community:
- a framework for allowing Linux distributions to pool their userbase and work together as one platform instead of multiple platforms, one per distro
- test suites which identifies both compatibility problems and outright bugs to be detected and fixed
- a method for targeting builds at multiple distributions at once, both proprietary and free
- reporting tools for finding portability problems in built apps, again for both proprietary and free apps
We currently provide all of this for 7 architectures. ARM benefits indirectly (for example, many of the compatibility breaks detected on, say, x86_64 will affect all archs equally), but indirect support doesn't include the tools we've developed, and often compatibility issues are arch-specific.
linaro-dev mailing list linaro-dev@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-dev
On Wed, Jun 1, 2011 at 1:25 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Tue, May 31, 2011 at 5:22 PM, Wookey wookey@wookware.org wrote:
In my experience anyone distributing binaries actually picks a small set of distros and builds for those explicitly, rather than relying on the LSB. Does that mean that it's not actually useful in the real world? I guess in a sense this posting is to the wrong lists; we're all free software people here who have little use for the LSB. Where do the proprietary software distributors hang out :-)
They are starting to hang out in all the familiar Free Software places. :-)
sooo... although the situation *right now* is that nobody in the commercial world is the slightest bit interested in LSB because they all do "custom builds" of complete software stacks, it could be said that *if* the free software community just dropped ready-to-go LSB standards in front of their noses, they'd quite likely use it.
Circumspect and balanced as always Mr. Leighton. :)
I've been to the "commercial world" on a temporary visa and they do in fact care about things like standards. In fact I would go so far as to say that this LSB proposal for ARM would significantly improve life for consortia like GENIVI which has members from the ARM and Intel camp.
you have to remember that the majority of these companies could not put two lines of code together to save their lives. they literally have to be spoon-fed (in some cases even to the point of being told where to put the screws, let alone the software). they are usually spoon-fed by the CPU manufacturer [and in the case of MStar Semi, they won't even let *you* violate the GPL, they do it entirely for you].
so in that regard, i think it's more a case of "if the free software community provides LSB across ARM, it'll get used".
I agree.
so in _that_ regard, the question becomes: "are the efforts of the free software community better off being spent elsewhere"? and "what benefit is there *TO THE FREE SOFTWARE COMMUNITY* of doing LSB for ARM"? forget the proprietary junkies, they'll suck anything from us that moves and not give a dime in return.
In my experience there are dimes to be had, you just have to ask nicely.
Regards,
Jeremiah
On Wed, 2011-06-01 at 12:25 +0100, Luke Kenneth Casson Leighton wrote:
sooo... although the situation *right now* is that nobody in the commercial world is the slightest bit interested in LSB because they all do "custom builds" of complete software stacks, it could be said that *if* the free software community just dropped ready-to-go LSB standards in front of their noses, they'd quite likely use it.
The reason we're discussing this is because a new architecture isn't going to be supported in standards like LSB overnight. It might take some time, and by that time, things may have changed with respect to the adoption of ARM systems. But if we don't think ahead, we're forced to be reactionary and try to do this (probably less effectively) later on.
Nobody will be forced to adopt LSB, but general purpose distributions can benefit from having compatibility at the software level. Is this an issue for deeply embedded platforms? Not so much. Is it bad that Android rebuilds the Universe? It's their decision to make. I think we need to distinguish between traditional embedded uses of ARM parts, which will continue, and those of larger parts running general a purpose OS. I don't expect to see Fedora running on my cellphone, but I do have it running on a netbook quite nicely - the latter needs LSB more.
I'll leave the rest of the rhetoric alone :)
Jon.