Can we deprecate ioctl(RAID_VERSION)?

Can we deprecate ioctl(RAID_VERSION)?

[jes.sorensen]

Hi Neil,

Looking through the code in mdadm, I noticed a number of cases calling
ioctl(RAID_VERSION). At first I had it confused with metadata version,
but it looks like RAID_VERSION will always return 90000 if it's a valid
raid device.

In the cases we want to confirm the fd is a valid raid array,
ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).

Am I missing something obvious here, or do you see any reason for
leaving this around?

Thanks,
Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[jes.sorensen]

jes.sorensen@gmail.com writes:
> Hi Neil,
>
> Looking through the code in mdadm, I noticed a number of cases calling
> ioctl(RAID_VERSION). At first I had it confused with metadata version,
> but it looks like RAID_VERSION will always return 90000 if it's a valid
> raid device.
>
> In the cases we want to confirm the fd is a valid raid array,
> ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).
>
> Am I missing something obvious here, or do you see any reason for
> leaving this around?

Sorry the above is wrong, it will always return 900, not 90000. Some of
the code that stood out is in util.c:

int md_get_version(int fd)
{
        struct stat stb;
        mdu_version_t vers;

        if (fstat(fd, &stb)<0)
                return -1;
        if ((S_IFMT&stb.st_mode) != S_IFBLK)
                return -1;

        if (ioctl(fd, RAID_VERSION, &vers) == 0)
                return  (vers.major*10000) + (vers.minor*100) + vers.patchlevel;
        if (errno == EACCES)
                return -1;
        if (major(stb.st_rdev) == MD_MAJOR)
                return (3600);
        return -1;
}

....

int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
{
        /* Initialise kernel's knowledge of array.
         * This varies between externally managed arrays
         * and older kernels
         */
        int vers = md_get_version(mdfd);
        int rv;

#ifndef MDASSEMBLE
        if (st->ss->external)
                rv = sysfs_set_array(info, vers);
        else
#endif
                if ((vers % 100) >= 1) { /* can use different versions */
                mdu_array_info_t inf;
                memset(&inf, 0, sizeof(inf));
                inf.major_version = info->array.major_version;
                inf.minor_version = info->array.minor_version;
                rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
        } else
                rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
        return rv;
}

This has been around since at least 2008, the current code came in
f35f25259279573c6274e2783536c0b0a399bdd4, but it looks like even the
prior code made the same assumptions.

In either case, the above 'if ((vers % 100) >= 1)' will always trigger
since the kernel does #define MD_PATCHLEVEL_VERSION 3

It's not like we have been updating MD_PATCHLEVEL_VERSION for a
while. Was the code meant to be looking at the superblock minor version?
I've been staring at this for a while now, so please beat me over the
head if I missed something blatantly obvious.

Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[NeilBrown]

[-- Attachment #1: Type: text/plain, Size: 3647 bytes --]

On Thu, Apr 06 2017, jes.sorensen@gmail.com wrote:

> jes.sorensen@gmail.com writes:
>> Hi Neil,
>>
>> Looking through the code in mdadm, I noticed a number of cases calling
>> ioctl(RAID_VERSION). At first I had it confused with metadata version,
>> but it looks like RAID_VERSION will always return 90000 if it's a valid
>> raid device.
>>
>> In the cases we want to confirm the fd is a valid raid array,
>> ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).
>>
>> Am I missing something obvious here, or do you see any reason for
>> leaving this around?
>
> Sorry the above is wrong, it will always return 900, not 90000. Some of
> the code that stood out is in util.c:
>
> int md_get_version(int fd)
> {
>         struct stat stb;
>         mdu_version_t vers;
>
>         if (fstat(fd, &stb)<0)
>                 return -1;
>         if ((S_IFMT&stb.st_mode) != S_IFBLK)
>                 return -1;
>
>         if (ioctl(fd, RAID_VERSION, &vers) == 0)
>                 return  (vers.major*10000) + (vers.minor*100) + vers.patchlevel;
>         if (errno == EACCES)
>                 return -1;
>         if (major(stb.st_rdev) == MD_MAJOR)
>                 return (3600);
>         return -1;
> }
>
> ....
>
> int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
> {
>         /* Initialise kernel's knowledge of array.
>          * This varies between externally managed arrays
>          * and older kernels
>          */
>         int vers = md_get_version(mdfd);
>         int rv;
>
> #ifndef MDASSEMBLE
>         if (st->ss->external)
>                 rv = sysfs_set_array(info, vers);
>         else
> #endif
>                 if ((vers % 100) >= 1) { /* can use different versions */
>                 mdu_array_info_t inf;
>                 memset(&inf, 0, sizeof(inf));
>                 inf.major_version = info->array.major_version;
>                 inf.minor_version = info->array.minor_version;
>                 rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
>         } else
>                 rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
>         return rv;
> }
>
> This has been around since at least 2008, the current code came in
> f35f25259279573c6274e2783536c0b0a399bdd4, but it looks like even the
> prior code made the same assumptions.
>
> In either case, the above 'if ((vers % 100) >= 1)' will always trigger
> since the kernel does #define MD_PATCHLEVEL_VERSION 3
>
> It's not like we have been updating MD_PATCHLEVEL_VERSION for a
> while. Was the code meant to be looking at the superblock minor version?
> I've been staring at this for a while now, so please beat me over the
> head if I missed something blatantly obvious.
>
> Jes

It is hard to get versioning right...

The version returned by the RAID_VERSION ioctl is meant to reflect the
capabilities of the implementation.  We could use the kernel version
number for that (and sometimes do), but as distro's often backport
features, that isn't always reliable.

I've incremented the MD_PATCHLEVEL_VERSION when a change is made that
cannot easily be detected from user-space.  As you note, we are up to
three.  The last change was in 2.6.15.
I've never contemplated changing the other two numbers that RAID_VERSION
return.  They don't seem to mean anything useful.

What exactly do you mean by "deprecate" the ioctl?
If you remove the code in mdadm that calls it, mdadm will not work
correctly on kernels older than 2.6.15, and it will be harder to
and an future capability that is not easily visible from user space.
If you remove the code in the kernel that handles it, you'll break
mdadm.

What is the goal here?

Thanks,
NeilBrown

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Can we deprecate ioctl(RAID_VERSION)?

[Jes Sorensen]

On 04/05/2017 06:32 PM, NeilBrown wrote:
> On Thu, Apr 06 2017, jes.sorensen@gmail.com wrote:
>
>> jes.sorensen@gmail.com writes:
>>> Hi Neil,
>>>
>>> Looking through the code in mdadm, I noticed a number of cases calling
>>> ioctl(RAID_VERSION). At first I had it confused with metadata version,
>>> but it looks like RAID_VERSION will always return 90000 if it's a valid
>>> raid device.
>>>
>>> In the cases we want to confirm the fd is a valid raid array,
>>> ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).
>>>
>>> Am I missing something obvious here, or do you see any reason for
>>> leaving this around?
>>
>> Sorry the above is wrong, it will always return 900, not 90000. Some of
>> the code that stood out is in util.c:
>>
>> int md_get_version(int fd)
>> {
>>         struct stat stb;
>>         mdu_version_t vers;
>>
>>         if (fstat(fd, &stb)<0)
>>                 return -1;
>>         if ((S_IFMT&stb.st_mode) != S_IFBLK)
>>                 return -1;
>>
>>         if (ioctl(fd, RAID_VERSION, &vers) == 0)
>>                 return  (vers.major*10000) + (vers.minor*100) + vers.patchlevel;
>>         if (errno == EACCES)
>>                 return -1;
>>         if (major(stb.st_rdev) == MD_MAJOR)
>>                 return (3600);
>>         return -1;
>> }
>>
>> ....
>>
>> int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
>> {
>>         /* Initialise kernel's knowledge of array.
>>          * This varies between externally managed arrays
>>          * and older kernels
>>          */
>>         int vers = md_get_version(mdfd);
>>         int rv;
>>
>> #ifndef MDASSEMBLE
>>         if (st->ss->external)
>>                 rv = sysfs_set_array(info, vers);
>>         else
>> #endif
>>                 if ((vers % 100) >= 1) { /* can use different versions */
>>                 mdu_array_info_t inf;
>>                 memset(&inf, 0, sizeof(inf));
>>                 inf.major_version = info->array.major_version;
>>                 inf.minor_version = info->array.minor_version;
>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
>>         } else
>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
>>         return rv;
>> }
>>
>> This has been around since at least 2008, the current code came in
>> f35f25259279573c6274e2783536c0b0a399bdd4, but it looks like even the
>> prior code made the same assumptions.
>>
>> In either case, the above 'if ((vers % 100) >= 1)' will always trigger
>> since the kernel does #define MD_PATCHLEVEL_VERSION 3
>>
>> It's not like we have been updating MD_PATCHLEVEL_VERSION for a
>> while. Was the code meant to be looking at the superblock minor version?
>> I've been staring at this for a while now, so please beat me over the
>> head if I missed something blatantly obvious.
>>
>> Jes
>
> It is hard to get versioning right...
>
> The version returned by the RAID_VERSION ioctl is meant to reflect the
> capabilities of the implementation.  We could use the kernel version
> number for that (and sometimes do), but as distro's often backport
> features, that isn't always reliable.
>
> I've incremented the MD_PATCHLEVEL_VERSION when a change is made that
> cannot easily be detected from user-space.  As you note, we are up to
> three.  The last change was in 2.6.15.
> I've never contemplated changing the other two numbers that RAID_VERSION
> return.  They don't seem to mean anything useful.
>
> What exactly do you mean by "deprecate" the ioctl?
> If you remove the code in mdadm that calls it, mdadm will not work
> correctly on kernels older than 2.6.15, and it will be harder to
> and an future capability that is not easily visible from user space.
> If you remove the code in the kernel that handles it, you'll break
> mdadm.

Neil,

I see, thanks for explaining.

The goal is to eventually get out of the ioctl() business and get to a 
state where we can do everything via sysfs/configfs. Right now we have a 
big mix between ioctl and sysfs where neither interface does everything. 
The recent issues with PPL (I think it was) showed that we had to add 
more ioctl support because the interfaces needed to do it for sysfs 
weren't quite there. My long term goal is to get that situation improved 
so we can avoid adding anymore ioctl interfaces and eventually allow for 
distros to build mdadm with ioctl support disabled. We had a discussion 
at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).

Reading the code I found it confusing that it was so tied to the patch 
level, but didn't do anything with the version numbers. At least 
intuitively if I bumped the version number, I would reset the patchlevel 
which would break things.

I think it's fair to draw a line in the sand and say that mdadm-4.1+ 
will not support kernels older than 2.6.15. I am open to the kernel 
version we pick here, but I would like to start deprecating some of the 
really old code. I have patches that does this in my tree, but I need to 
add a check for kernel version > 2.6.15. I am not aware what SuSE's 
enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was 
2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite 
a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems fine.

For the kernel to expose features to userland in the future, I would 
prefer to go with a feature-flag style interface exposed via sysfs. That 
way a distro could enable one feature, but not the other in their kernel 
without having to worry about actual version numbers.

Cheers,
Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[Coly Li]

On 2017/4/6 下午11:31, Jes Sorensen wrote:
> On 04/05/2017 06:32 PM, NeilBrown wrote:
>> On Thu, Apr 06 2017, jes.sorensen@gmail.com wrote:
>>
>>> jes.sorensen@gmail.com writes:
>>>> Hi Neil,
>>>>
>>>> Looking through the code in mdadm, I noticed a number of cases calling
>>>> ioctl(RAID_VERSION). At first I had it confused with metadata version,
>>>> but it looks like RAID_VERSION will always return 90000 if it's a valid
>>>> raid device.
>>>>
>>>> In the cases we want to confirm the fd is a valid raid array,
>>>> ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).
>>>>
>>>> Am I missing something obvious here, or do you see any reason for
>>>> leaving this around?
>>>
>>> Sorry the above is wrong, it will always return 900, not 90000. Some of
>>> the code that stood out is in util.c:
>>>
>>> int md_get_version(int fd)
>>> {
>>>         struct stat stb;
>>>         mdu_version_t vers;
>>>
>>>         if (fstat(fd, &stb)<0)
>>>                 return -1;
>>>         if ((S_IFMT&stb.st_mode) != S_IFBLK)
>>>                 return -1;
>>>
>>>         if (ioctl(fd, RAID_VERSION, &vers) == 0)
>>>                 return  (vers.major*10000) + (vers.minor*100) +
>>> vers.patchlevel;
>>>         if (errno == EACCES)
>>>                 return -1;
>>>         if (major(stb.st_rdev) == MD_MAJOR)
>>>                 return (3600);
>>>         return -1;
>>> }
>>>
>>> ....
>>>
>>> int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
>>> {
>>>         /* Initialise kernel's knowledge of array.
>>>          * This varies between externally managed arrays
>>>          * and older kernels
>>>          */
>>>         int vers = md_get_version(mdfd);
>>>         int rv;
>>>
>>> #ifndef MDASSEMBLE
>>>         if (st->ss->external)
>>>                 rv = sysfs_set_array(info, vers);
>>>         else
>>> #endif
>>>                 if ((vers % 100) >= 1) { /* can use different
>>> versions */
>>>                 mdu_array_info_t inf;
>>>                 memset(&inf, 0, sizeof(inf));
>>>                 inf.major_version = info->array.major_version;
>>>                 inf.minor_version = info->array.minor_version;
>>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
>>>         } else
>>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
>>>         return rv;
>>> }
>>>
>>> This has been around since at least 2008, the current code came in
>>> f35f25259279573c6274e2783536c0b0a399bdd4, but it looks like even the
>>> prior code made the same assumptions.
>>>
>>> In either case, the above 'if ((vers % 100) >= 1)' will always trigger
>>> since the kernel does #define MD_PATCHLEVEL_VERSION 3
>>>
>>> It's not like we have been updating MD_PATCHLEVEL_VERSION for a
>>> while. Was the code meant to be looking at the superblock minor version?
>>> I've been staring at this for a while now, so please beat me over the
>>> head if I missed something blatantly obvious.
>>>
>>> Jes
>>
>> It is hard to get versioning right...
>>
>> The version returned by the RAID_VERSION ioctl is meant to reflect the
>> capabilities of the implementation.  We could use the kernel version
>> number for that (and sometimes do), but as distro's often backport
>> features, that isn't always reliable.
>>
>> I've incremented the MD_PATCHLEVEL_VERSION when a change is made that
>> cannot easily be detected from user-space.  As you note, we are up to
>> three.  The last change was in 2.6.15.
>> I've never contemplated changing the other two numbers that RAID_VERSION
>> return.  They don't seem to mean anything useful.
>>
>> What exactly do you mean by "deprecate" the ioctl?
>> If you remove the code in mdadm that calls it, mdadm will not work
>> correctly on kernels older than 2.6.15, and it will be harder to
>> and an future capability that is not easily visible from user space.
>> If you remove the code in the kernel that handles it, you'll break
>> mdadm.
> 
> Neil,
> 
> I see, thanks for explaining.
> 
> The goal is to eventually get out of the ioctl() business and get to a
> state where we can do everything via sysfs/configfs. Right now we have a
> big mix between ioctl and sysfs where neither interface does everything.
> The recent issues with PPL (I think it was) showed that we had to add
> more ioctl support because the interfaces needed to do it for sysfs
> weren't quite there. My long term goal is to get that situation improved
> so we can avoid adding anymore ioctl interfaces and eventually allow for
> distros to build mdadm with ioctl support disabled. We had a discussion
> at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).
> 
> Reading the code I found it confusing that it was so tied to the patch
> level, but didn't do anything with the version numbers. At least
> intuitively if I bumped the version number, I would reset the patchlevel
> which would break things.
> 
> I think it's fair to draw a line in the sand and say that mdadm-4.1+
> will not support kernels older than 2.6.15. I am open to the kernel
> version we pick here, but I would like to start deprecating some of the
> really old code. I have patches that does this in my tree, but I need to
> add a check for kernel version > 2.6.15. I am not aware what SuSE's
> enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was
> 2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite
> a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems
> fine.
> 
> For the kernel to expose features to userland in the future, I would
> prefer to go with a feature-flag style interface exposed via sysfs. That
> way a distro could enable one feature, but not the other in their kernel
> without having to worry about actual version numbers.

BTW, I'd like to take on the kernel side stuffs.
IMHO, it would be better that we also set a timeline, after the
timeline, we should add new interface via configfs, and keep all legancy
ioctl implementation before this timeline.

Coly

Re: Can we deprecate ioctl(RAID_VERSION)?

[Jes Sorensen]

On 04/06/2017 02:10 PM, Coly Li wrote:
> On 2017/4/6 下午11:31, Jes Sorensen wrote:
>> Neil,
>>
>> I see, thanks for explaining.
>>
>> The goal is to eventually get out of the ioctl() business and get to a
>> state where we can do everything via sysfs/configfs. Right now we have a
>> big mix between ioctl and sysfs where neither interface does everything.
>> The recent issues with PPL (I think it was) showed that we had to add
>> more ioctl support because the interfaces needed to do it for sysfs
>> weren't quite there. My long term goal is to get that situation improved
>> so we can avoid adding anymore ioctl interfaces and eventually allow for
>> distros to build mdadm with ioctl support disabled. We had a discussion
>> at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).
>>
>> Reading the code I found it confusing that it was so tied to the patch
>> level, but didn't do anything with the version numbers. At least
>> intuitively if I bumped the version number, I would reset the patchlevel
>> which would break things.
>>
>> I think it's fair to draw a line in the sand and say that mdadm-4.1+
>> will not support kernels older than 2.6.15. I am open to the kernel
>> version we pick here, but I would like to start deprecating some of the
>> really old code. I have patches that does this in my tree, but I need to
>> add a check for kernel version > 2.6.15. I am not aware what SuSE's
>> enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was
>> 2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite
>> a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems
>> fine.
>>
>> For the kernel to expose features to userland in the future, I would
>> prefer to go with a feature-flag style interface exposed via sysfs. That
>> way a distro could enable one feature, but not the other in their kernel
>> without having to worry about actual version numbers.
>
> BTW, I'd like to take on the kernel side stuffs.
> IMHO, it would be better that we also set a timeline, after the
> timeline, we should add new interface via configfs, and keep all legancy
> ioctl implementation before this timeline.
>
> Coly

A volunteer! A volunteer!

I don't think we need to set a firm timeline for removing code from the 
kernel at this point. What we can do is implement the new interfaces via 
sysfs/configfs, and then make ioctl support a config option. Down the 
line it will be evident if/when we can rip out the old code, but I see 
that being years away.

Cheers,
Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[Coly Li]

On 2017/4/7 上午2:14, Jes Sorensen wrote:
> On 04/06/2017 02:10 PM, Coly Li wrote:
>> On 2017/4/6 下午11:31, Jes Sorensen wrote:
>>> Neil,
>>>
>>> I see, thanks for explaining.
>>>
>>> The goal is to eventually get out of the ioctl() business and get to a
>>> state where we can do everything via sysfs/configfs. Right now we have a
>>> big mix between ioctl and sysfs where neither interface does everything.
>>> The recent issues with PPL (I think it was) showed that we had to add
>>> more ioctl support because the interfaces needed to do it for sysfs
>>> weren't quite there. My long term goal is to get that situation improved
>>> so we can avoid adding anymore ioctl interfaces and eventually allow for
>>> distros to build mdadm with ioctl support disabled. We had a discussion
>>> at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).
>>>
>>> Reading the code I found it confusing that it was so tied to the patch
>>> level, but didn't do anything with the version numbers. At least
>>> intuitively if I bumped the version number, I would reset the patchlevel
>>> which would break things.
>>>
>>> I think it's fair to draw a line in the sand and say that mdadm-4.1+
>>> will not support kernels older than 2.6.15. I am open to the kernel
>>> version we pick here, but I would like to start deprecating some of the
>>> really old code. I have patches that does this in my tree, but I need to
>>> add a check for kernel version > 2.6.15. I am not aware what SuSE's
>>> enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was
>>> 2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite
>>> a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems
>>> fine.
>>>
>>> For the kernel to expose features to userland in the future, I would
>>> prefer to go with a feature-flag style interface exposed via sysfs. That
>>> way a distro could enable one feature, but not the other in their kernel
>>> without having to worry about actual version numbers.
>>
>> BTW, I'd like to take on the kernel side stuffs.
>> IMHO, it would be better that we also set a timeline, after the
>> timeline, we should add new interface via configfs, and keep all legancy
>> ioctl implementation before this timeline.
>>
>> Coly
> 
> A volunteer! A volunteer!
> 
> I don't think we need to set a firm timeline for removing code from the
> kernel at this point. What we can do is implement the new interfaces via
> sysfs/configfs, and then make ioctl support a config option. Down the
> line it will be evident if/when we can rip out the old code, but I see
> that being years away.

Not removing code, just after a timeline, new patch which adding new
interface should go into configfs.
I start to look into this now, hope to compose a first version patch set
for comments in not too long time.

Coly

Re: Can we deprecate ioctl(RAID_VERSION)?

[Jes Sorensen]

On 04/06/2017 11:54 PM, Coly Li wrote:
> On 2017/4/7 上午2:14, Jes Sorensen wrote:
>> A volunteer! A volunteer!
>>
>> I don't think we need to set a firm timeline for removing code from the
>> kernel at this point. What we can do is implement the new interfaces via
>> sysfs/configfs, and then make ioctl support a config option. Down the
>> line it will be evident if/when we can rip out the old code, but I see
>> that being years away.
>
> Not removing code, just after a timeline, new patch which adding new
> interface should go into configfs.
> I start to look into this now, hope to compose a first version patch set
> for comments in not too long time.
>
> Coly

Sounds good - make sure to also consult with Shaohua how it likes to see 
it implemented on the kernel side.

Cheers,
Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[NeilBrown]

[-- Attachment #1: Type: text/plain, Size: 6619 bytes --]

On Thu, Apr 06 2017, Jes Sorensen wrote:

> On 04/05/2017 06:32 PM, NeilBrown wrote:
>> On Thu, Apr 06 2017, jes.sorensen@gmail.com wrote:
>>
>>> jes.sorensen@gmail.com writes:
>>>> Hi Neil,
>>>>
>>>> Looking through the code in mdadm, I noticed a number of cases calling
>>>> ioctl(RAID_VERSION). At first I had it confused with metadata version,
>>>> but it looks like RAID_VERSION will always return 90000 if it's a valid
>>>> raid device.
>>>>
>>>> In the cases we want to confirm the fd is a valid raid array,
>>>> ioctl(GET_ARRAY_INFO) should do, or sysfs_read(GET_VERSION).
>>>>
>>>> Am I missing something obvious here, or do you see any reason for
>>>> leaving this around?
>>>
>>> Sorry the above is wrong, it will always return 900, not 90000. Some of
>>> the code that stood out is in util.c:
>>>
>>> int md_get_version(int fd)
>>> {
>>>         struct stat stb;
>>>         mdu_version_t vers;
>>>
>>>         if (fstat(fd, &stb)<0)
>>>                 return -1;
>>>         if ((S_IFMT&stb.st_mode) != S_IFBLK)
>>>                 return -1;
>>>
>>>         if (ioctl(fd, RAID_VERSION, &vers) == 0)
>>>                 return  (vers.major*10000) + (vers.minor*100) + vers.patchlevel;
>>>         if (errno == EACCES)
>>>                 return -1;
>>>         if (major(stb.st_rdev) == MD_MAJOR)
>>>                 return (3600);
>>>         return -1;
>>> }
>>>
>>> ....
>>>
>>> int set_array_info(int mdfd, struct supertype *st, struct mdinfo *info)
>>> {
>>>         /* Initialise kernel's knowledge of array.
>>>          * This varies between externally managed arrays
>>>          * and older kernels
>>>          */
>>>         int vers = md_get_version(mdfd);
>>>         int rv;
>>>
>>> #ifndef MDASSEMBLE
>>>         if (st->ss->external)
>>>                 rv = sysfs_set_array(info, vers);
>>>         else
>>> #endif
>>>                 if ((vers % 100) >= 1) { /* can use different versions */
>>>                 mdu_array_info_t inf;
>>>                 memset(&inf, 0, sizeof(inf));
>>>                 inf.major_version = info->array.major_version;
>>>                 inf.minor_version = info->array.minor_version;
>>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, &inf);
>>>         } else
>>>                 rv = ioctl(mdfd, SET_ARRAY_INFO, NULL);
>>>         return rv;
>>> }
>>>
>>> This has been around since at least 2008, the current code came in
>>> f35f25259279573c6274e2783536c0b0a399bdd4, but it looks like even the
>>> prior code made the same assumptions.
>>>
>>> In either case, the above 'if ((vers % 100) >= 1)' will always trigger
>>> since the kernel does #define MD_PATCHLEVEL_VERSION 3
>>>
>>> It's not like we have been updating MD_PATCHLEVEL_VERSION for a
>>> while. Was the code meant to be looking at the superblock minor version?
>>> I've been staring at this for a while now, so please beat me over the
>>> head if I missed something blatantly obvious.
>>>
>>> Jes
>>
>> It is hard to get versioning right...
>>
>> The version returned by the RAID_VERSION ioctl is meant to reflect the
>> capabilities of the implementation.  We could use the kernel version
>> number for that (and sometimes do), but as distro's often backport
>> features, that isn't always reliable.
>>
>> I've incremented the MD_PATCHLEVEL_VERSION when a change is made that
>> cannot easily be detected from user-space.  As you note, we are up to
>> three.  The last change was in 2.6.15.
>> I've never contemplated changing the other two numbers that RAID_VERSION
>> return.  They don't seem to mean anything useful.
>>
>> What exactly do you mean by "deprecate" the ioctl?
>> If you remove the code in mdadm that calls it, mdadm will not work
>> correctly on kernels older than 2.6.15, and it will be harder to
>> and an future capability that is not easily visible from user space.
>> If you remove the code in the kernel that handles it, you'll break
>> mdadm.
>
> Neil,
>
> I see, thanks for explaining.
>
> The goal is to eventually get out of the ioctl() business and get to a 
> state where we can do everything via sysfs/configfs. Right now we have a 
> big mix between ioctl and sysfs where neither interface does everything. 
> The recent issues with PPL (I think it was) showed that we had to add 
> more ioctl support because the interfaces needed to do it for sysfs 
> weren't quite there. My long term goal is to get that situation improved 
> so we can avoid adding anymore ioctl interfaces and eventually allow for 
> distros to build mdadm with ioctl support disabled. We had a discussion 
> at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).

Sounds like a good goal, if approached cautiously (and it does sound
like you are showing proper caution).
And I don't think we need the "/configfs" bit.  Configfs is just for
people who don't understand sysfs ;-)

>
> Reading the code I found it confusing that it was so tied to the patch 
> level, but didn't do anything with the version numbers. At least 
> intuitively if I bumped the version number, I would reset the patchlevel 
> which would break things.

I assumed the version number would never change, because it didn't mean
anything useful.

>
> I think it's fair to draw a line in the sand and say that mdadm-4.1+ 
> will not support kernels older than 2.6.15. I am open to the kernel 
> version we pick here, but I would like to start deprecating some of the 
> really old code. I have patches that does this in my tree, but I need to 
> add a check for kernel version > 2.6.15. I am not aware what SuSE's 
> enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was 
> 2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite 
> a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems fine.

With my SUSE hat on, I'm happy for new mdadm to not support kernels
older than 3.0.  Probably even 3.12.

>
> For the kernel to expose features to userland in the future, I would 
> prefer to go with a feature-flag style interface exposed via sysfs. That 
> way a distro could enable one feature, but not the other in their kernel 
> without having to worry about actual version numbers.

I think there are usually better ways than feature flags.
If the new feature requires a new file in sysfs, then the existence of
the file signals the presence of the feature.

I'm happy with a plan to freeze the patchlevel at its current value and
use other mechanisms going forward.

Thanks,
NeilBrown

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Can we deprecate ioctl(RAID_VERSION)?

[jes.sorensen]

NeilBrown <neilb@suse.com> writes:
> On Thu, Apr 06 2017, Jes Sorensen wrote:
>> Neil,
>>
>> I see, thanks for explaining.
>>
>> The goal is to eventually get out of the ioctl() business and get to a 
>> state where we can do everything via sysfs/configfs. Right now we have a 
>> big mix between ioctl and sysfs where neither interface does everything. 
>> The recent issues with PPL (I think it was) showed that we had to add 
>> more ioctl support because the interfaces needed to do it for sysfs 
>> weren't quite there. My long term goal is to get that situation improved 
>> so we can avoid adding anymore ioctl interfaces and eventually allow for 
>> distros to build mdadm with ioctl support disabled. We had a discussion 
>> at LSF/MM in Boston about this (Hannes, Shaohua, Song, and myself).
>
> Sounds like a good goal, if approached cautiously (and it does sound
> like you are showing proper caution).
> And I don't think we need the "/configfs" bit.  Configfs is just for
> people who don't understand sysfs ;-)

Glad to hear we're in agreement on this. I definitely want to be
cautious about this. While change is good, we have a responsibility
towards existing users. This isn't a desktop environment after all :)

I am not really tied to sysfs/configfs on this, whoever does that part
of the work gets to show us what he/she works best and explain why.

>> I think it's fair to draw a line in the sand and say that mdadm-4.1+ 
>> will not support kernels older than 2.6.15. I am open to the kernel 
>> version we pick here, but I would like to start deprecating some of the 
>> really old code. I have patches that does this in my tree, but I need to 
>> add a check for kernel version > 2.6.15. I am not aware what SuSE's 
>> enterprise kernel versions look like, but checking RHEL/CentOS RHEL5 was 
>> 2.6.18, while RHEL4 was 2.6.9 - and RHEL4 has been unsupported for quite 
>> a while. At least for RHEL/CentOS 2.6.15 as the line in the sand seems fine.
>
> With my SUSE hat on, I'm happy for new mdadm to not support kernels
> older than 3.0.  Probably even 3.12.

I just pushed the first set of changes into git for this. We no longer
support kernels older than 2.6.15.

If it breaks something else, I am ready to take public blame! If it ends
up biting another distro for a slightly older kernel, we can look at
fixing that up.

>> For the kernel to expose features to userland in the future, I would 
>> prefer to go with a feature-flag style interface exposed via sysfs. That 
>> way a distro could enable one feature, but not the other in their kernel 
>> without having to worry about actual version numbers.
>
> I think there are usually better ways than feature flags.
> If the new feature requires a new file in sysfs, then the existence of
> the file signals the presence of the feature.

That is pretty much how I see feature flags.

Next question since I am wearing my 'what is this old stuff doing' hat.
mdassemble? Does anything still use this? The reason is a lot of the
newer features are explicitly included, and switching to sysfs is
effectively going to kill it, unless it gets a major upgrade.

Cheers,
Jes

Re: Can we deprecate ioctl(RAID_VERSION)?

[NeilBrown]

[-- Attachment #1: Type: text/plain, Size: 583 bytes --]

On Fri, Apr 07 2017, jes.sorensen@gmail.com wrote:

>
> Next question since I am wearing my 'what is this old stuff doing' hat.
> mdassemble? Does anything still use this? The reason is a lot of the
> newer features are explicitly included, and switching to sysfs is
> effectively going to kill it, unless it gets a major upgrade.
>

I was never a big fan, of mdassemble, but it is smaller than mdadm and
some people apparently have (or had) space-constrained boot
environments.
Maybe post to the linux-raid list with a subject "mdassemble is going
way, do you care?". ??

NeilBrown

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Can we deprecate ioctl(RAID_VERSION)?

[Nix]

On 10 Apr 2017, NeilBrown verbalised:

> On Fri, Apr 07 2017, jes.sorensen@gmail.com wrote:
>
>> Next question since I am wearing my 'what is this old stuff doing' hat.
>> mdassemble? Does anything still use this? The reason is a lot of the
>> newer features are explicitly included, and switching to sysfs is
>> effectively going to kill it, unless it gets a major upgrade.
>>
>
> I was never a big fan, of mdassemble, but it is smaller than mdadm and
> some people apparently have (or had) space-constrained boot
> environments.

It also has fewer build-time requirements and can build on systems with
things like old buggy versions of uclibc that can't build a working
mdadm. (Of course, now musl exists, I'm not sure anyone should care
about that. I stopped caring many years ago.)

-- 
NULL && (void)

Re: Can we deprecate ioctl(RAID_VERSION)?

[Jes Sorensen]

On 04/09/2017 07:01 PM, NeilBrown wrote:
> On Fri, Apr 07 2017, jes.sorensen@gmail.com wrote:
>
>>
>> Next question since I am wearing my 'what is this old stuff doing' hat.
>> mdassemble? Does anything still use this? The reason is a lot of the
>> newer features are explicitly included, and switching to sysfs is
>> effectively going to kill it, unless it gets a major upgrade.
>>
>
> I was never a big fan, of mdassemble, but it is smaller than mdadm and
> some people apparently have (or had) space-constrained boot
> environments.
> Maybe post to the linux-raid list with a subject "mdassemble is going
> way, do you care?". ??

Makes sense - looking at the code it wasn't clear if it is even usable 
in modern times since it ignores sysfs access.

I'll post another message and get out the big hammer!

Cheers,
Jes