22bb1be4d27... breaks hal (and NM) - Kernel

This is a discussion on 22bb1be4d27... breaks hal (and NM) - Kernel ; From commit 22bb1be4d27...: "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used by anyone." Can't have been much of an investigation as HAL uses these to determine if a card is a wireless one. Without this option, you ...

+ Reply to Thread
Results 1 to 10 of 10

Thread: 22bb1be4d27... breaks hal (and NM)

  1. 22bb1be4d27... breaks hal (and NM)

    From commit 22bb1be4d27...:

    "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    by anyone."

    Can't have been much of an investigation as HAL uses these to determine
    if a card is a wireless one. Without this option, you cannot get
    wireless support in NetworkManager, stranding most laptops.

    Change the default and mark it for future deprecation if you want to
    get rid of it.

    --
    -- Pierre Ossman

    Linux kernel, MMC maintainer http://www.kernel.org
    rdesktop, core developer http://www.rdesktop.org

    WARNING: This correspondence is being monitored by the
    Swedish government. Make sure your server uses encryption
    for SMTP traffic and consider using PGP for end-to-end
    encryption.

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v2.0.9 (GNU/Linux)

    iEYEARECAAYFAkiIUjAACgkQ7b8eESbyJLgfYgCguKna1rLwwW UFArNTpHzkYwEE
    wWYAoIW6oLc7pE/ulaq1RKr63BLjbI5M
    =j/he
    -----END PGP SIGNATURE-----


  2. Re: 22bb1be4d27... breaks hal (and NM)

    On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > From commit 22bb1be4d27...:
    >
    > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > by anyone."
    >
    > Can't have been much of an investigation as HAL uses these to determine
    > if a card is a wireless one. Without this option, you cannot get
    > wireless support in NetworkManager, stranding most laptops.


    Bzzzzt. You suck. Go check again. Hint: try hal's git tree.

    johannes

    -----BEGIN PGP SIGNATURE-----
    Comment: Johannes Berg (powerbook)

    iQIcBAABAgAGBQJIiGm9AAoJEKVg1VMiehFYQIwQAJ+0xPEhYn cyNSx29+Vq1nfy
    5L1HDmIbbUdOW8h7Ca54h5EqRaeTwV7DMEJLse3ca4x+Dtzigg fahL+yucDEE+I8
    MK6wQvbJ65gI/n7GzVB2/lbbc5NEVi2ADaDsWPRXXPgSbsT0do8VU69FnXyDpWqg
    NxmXau7bm5h801IfTyYRDrTqRyiP1bQ5QZd0XUbRxlyGYU44dX K31FjgtlD3Yi0x
    FQHmgAswTiAMlbIbLmkNCf1Wp+AsZicU3+N6+RibqwPHi8AGd8 pTFpxnjbGikRQj
    XjIFz79LG9dDtQzOEfFKYfZfWu8l/tMhHldrU7vhmBISEfAJAEupWZ7sXqCGfGdb
    zByp2jnePI7plnYWQfTL/aY/BRTmBx+Az+bDHxhtcNOSm2AJXkiRxbfZHr7mskUx
    XciT56ZOtGSeORGLFS0jjcK7ETOxs9sSR/n2x75IolEiB/+GXyoqQj4a+T1FtlEW
    yb/S4jVLxACCVk4vKxBpjLG7UXSaGK6G7EAiZ6cf3grMoKFjNYsEH 28dAjBkKEvu
    sipvxaY0d0RwlI97LXgmLk/k2QfP2IvwcUo625n9QAq6o4qLhmYBc756haDedlqF
    rYmIoo3gXAe75v98LC2G8OJB3gLdk+wrZHyKQtN3nxnITwutgS AaDwvecWGQs5Qa
    wVwC7/E7q27GdKpSElrv
    =YfaU
    -----END PGP SIGNATURE-----


  3. Re: 22bb1be4d27... breaks hal (and NM)

    On Thu, 2008-07-24 at 13:38 +0200, Johannes Berg wrote:
    > On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > > From commit 22bb1be4d27...:
    > >
    > > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > > by anyone."
    > >
    > > Can't have been much of an investigation as HAL uses these to determine
    > > if a card is a wireless one. Without this option, you cannot get
    > > wireless support in NetworkManager, stranding most laptops.

    >
    > Bzzzzt. You suck. Go check again. Hint: try hal's git tree.


    So you're not considering HAL as shipped by most (if not all) distros as
    anyone?

    If everybody would act like that, the next kernel wouldn't even manage
    to get rudimentary user-space up and running. Like we already changed
    all those systemcalls in glibc-head, who needs this kernel to boot on
    your old stuff anyway..

    Common,.. this is rediculous.

    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  4. Re: 22bb1be4d27... breaks hal (and NM)

    On Thu, 2008-07-24 at 13:57 +0200, Peter Zijlstra wrote:

    > So you're not considering HAL as shipped by most (if not all) distros as
    > anyone?
    >
    > If everybody would act like that, the next kernel wouldn't even manage
    > to get rudimentary user-space up and running. Like we already changed
    > all those systemcalls in glibc-head, who needs this kernel to boot on
    > your old stuff anyway..
    >
    > Common,.. this is rediculous.


    What the hell is wrong with you guys? I used to think it's accepted
    practice to introduce a config option you can change away from the
    default if you already know you are using new software that doesn't need
    it, and if you are clueless just don't ****ing touch the default until I
    decide it's time to flip the default/remove the option because distros
    are shipping new versions of the dependent software.

    johannes

    -----BEGIN PGP SIGNATURE-----
    Comment: Johannes Berg (powerbook)

    iQIcBAABAgAGBQJIiHCmAAoJEKVg1VMiehFYpJMQAI+5+Llx7f 89zQDTqvPUWQnn
    bvf2h9OTncMHY1DhAQh6l4H20rn5se16blwtBBK/L2PC3F5LylMqm64GMB83bBgo
    wjnudYqBO2NpPCXwTUexubBFsiatuFKp21gU8dbgo+9F+vSgbg DxoHySOJhqLZ5r
    LLgfDyya7u8rjIUXoLFuyOQD4gTZFfDjZ0DLSymst8uVnHCj1X/Wmy8hmjvon8+0
    Do/eyJrf394pGJFPtkKpSwaj4uZBNIf4Z7eH3oj1pTP7e5Poiubme defRJ17ziU8
    /F2HrEy91E2YPkjBbckLSd73oUcxTO8tZR/gut2Zp+sU0q1CizQc82VYDU/n5agb
    pz+Q4icuQCb5zDTVIq4aVYRRfWKuZ9SyFQrc1CcrTKzgK8NITT DINJGuEoMASwJH
    u+5EPhMBzpENCoxc6QilJ35yZkd+xo3Wh5fQM3Wmv1gcuJLr4+ IH7uwUuCqL4Epp
    LfoBryfAgFYtfaHPHHeP8Qz9IegKPoDZNeYSLWnS9Yy4KgopYD eItmbeVX4XtuTy
    tKIa/rn+GSj99i6LCbWyRWBlYvtyTes1uE69CJ9smaaHITu5GjNs1Pk gUnGewbfE
    Agu3066BBqg41CzjygJl5C27ARyZxwpsqlgT/Bvk4yZwKxkWyj7fVoc0+g31UaF5
    Br5xFBXBhso1QfGjgL34
    =7AjK
    -----END PGP SIGNATURE-----


  5. Re: 22bb1be4d27... breaks hal (and NM)

    On Thu, 24 Jul 2008 13:57:11 +0200 Peter Zijlstra wrote:

    > On Thu, 2008-07-24 at 13:38 +0200, Johannes Berg wrote:
    > > On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > > > From commit 22bb1be4d27...:
    > > >
    > > > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > > > by anyone."
    > > >
    > > > Can't have been much of an investigation as HAL uses these to determine
    > > > if a card is a wireless one. Without this option, you cannot get
    > > > wireless support in NetworkManager, stranding most laptops.

    > >
    > > Bzzzzt. You suck. Go check again. Hint: try hal's git tree.


    err,

    a) don't be rude

    b) what's in various git trees is unuseful for making decisions
    about production kernel features.

    otoh the patch seems reasonable.

    > So you're not considering HAL as shipped by most (if not all) distros as
    > anyone?
    >
    > If everybody would act like that, the next kernel wouldn't even manage
    > to get rudimentary user-space up and running. Like we already changed
    > all those systemcalls in glibc-head, who needs this kernel to boot on
    > your old stuff anyway..
    >
    > Common,.. this is rediculous.


    I think you'd have needed to try pretty hard to let this break stuff.

    : commit 22bb1be4d271961846cd0889b0f8d671db773080
    : Author: Johannes Berg
    : Date: Thu Jul 10 11:16:47 2008 +0200
    :
    : wext: make sysfs bits optional and deprecate them
    :
    : The /sys/class/net/*/wireless/ direcory is, as far as I know, not
    : used by anyone. Additionally, the same data is available via wext
    : ioctls. Hence the sysfs files are pretty much useless. This patch
    : makes them optional and schedules them for removal.
    :
    : Signed-off-by: Johannes Berg
    : Cc: Jean Tourrilhes
    : Signed-off-by: John W. Linville
    :
    : diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt
    : index 8319c46..db300e0 100644
    : --- a/Documentation/feature-removal-schedule.txt
    : +++ b/Documentation/feature-removal-schedule.txt
    : @@ -333,3 +333,13 @@ Why: This option was introduced just to allow older lm-sensors userspace
    : to keep working over the upgrade to 2.6.26. At the scheduled time of
    : removal fixed lm-sensors (2.x or 3.x) should be readily available.
    : Who: Rene Herman
    : +
    : +---------------------------
    : +
    : +What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
    : + (in net/core/net-sysfs.c)
    : +When: After the only user (hal) has seen a release with the patches
    : + for enough time, probably some time in 2010.
    : +Why: Over 1K .text/.data size reduction, data is available in other
    : + ways (ioctls)
    : +Who: Johannes Berg
    : diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
    : index 3f79413..c1f4e0d 100644
    : --- a/net/core/net-sysfs.c
    : +++ b/net/core/net-sysfs.c
    : @@ -318,7 +318,7 @@ static struct attribute_group netstat_group = {
    : .attrs = netstat_attrs,
    : };
    :
    : -#ifdef CONFIG_WIRELESS_EXT
    : +#ifdef CONFIG_WIRELESS_EXT_SYSFS
    : /* helper function that does all the locking etc for wireless stats */
    : static ssize_t wireless_show(struct device *d, char *buf,
    : ssize_t (*format)(const struct iw_statistics *,
    : @@ -459,7 +459,7 @@ int netdev_register_kobject(struct net_device *net)
    : #ifdef CONFIG_SYSFS
    : *groups++ = &netstat_group;
    :
    : -#ifdef CONFIG_WIRELESS_EXT
    : +#ifdef CONFIG_WIRELESS_EXT_SYSFS
    : if (net->wireless_handlers && net->wireless_handlers->get_wireless_stats)
    : *groups++ = &wireless_group;
    : #endif
    : diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig
    : index 7927090..ab015c6 100644
    : --- a/net/wireless/Kconfig
    : +++ b/net/wireless/Kconfig
    : @@ -29,3 +29,14 @@ config WIRELESS_EXT
    :
    : Say N (if you can) unless you know you need wireless
    : extensions for external modules.
    : +
    : +config WIRELESS_EXT_SYSFS
    : + bool "Wireless extensions sysfs files"
    : + default y
    : + depends on WIRELESS_EXT && SYSFS
    : + help
    : + This option enables the deprecated wireless statistics
    : + files in /sys/class/net/*/wireless/. The same information
    : + is available via the ioctls as well.
    : +
    : + Say Y if you have programs using it (we don't know of any).
    :

    So if you've enabled CONFIG_WIRELESS_EXT_SYSFS (and it correctly has
    default y) then things should continue to work OK. And the deprecation
    date of 2010 sounds reasonable. In fact generous, for us...

    So where's the problem?
    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  6. Re: 22bb1be4d27... breaks hal (and NM)

    On Thu, 2008-07-24 at 05:11 -0700, Andrew Morton wrote:
    > On Thu, 24 Jul 2008 13:57:11 +0200 Peter Zijlstra wrote:
    >
    > > On Thu, 2008-07-24 at 13:38 +0200, Johannes Berg wrote:
    > > > On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > > > > From commit 22bb1be4d27...:
    > > > >
    > > > > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > > > > by anyone."
    > > > >
    > > > > Can't have been much of an investigation as HAL uses these to determine
    > > > > if a card is a wireless one. Without this option, you cannot get
    > > > > wireless support in NetworkManager, stranding most laptops.
    > > >
    > > > Bzzzzt. You suck. Go check again. Hint: try hal's git tree.

    >
    > err,
    >
    > a) don't be rude
    >
    > b) what's in various git trees is unuseful for making decisions
    > about production kernel features.
    >
    > otoh the patch seems reasonable.


    > So if you've enabled CONFIG_WIRELESS_EXT_SYSFS (and it correctly has
    > default y) then things should continue to work OK. And the deprecation
    > date of 2010 sounds reasonable. In fact generous, for us...
    >
    > So where's the problem?


    Your two previous points, a and b.

    That help text should really have mentioned current software, not future
    software. And the tone of the reply.

    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  7. Re: 22bb1be4d27... breaks hal (and NM)

    On Thursday 24 July 2008 13:11:13 Andrew Morton wrote:
    > : + Say Y if you have programs using it (we don't know of any).


    Probably would have been less confusing if the Kconfig message actually lined
    up with the note about older HALs in feature-removal-schedule.txt. Saying "we
    don't know of any" seems inaccurate. This should really be something
    like "old HAL versions" or perhaps even better "HAL <= X.Y.Z".

    I'm sure distributors will have the common sense to stop enabling the option
    in the future. In the mean time, removing the deceptive summary would
    hopefully avoid such bug reports.

    --
    Cheers,
    Alistair.
    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  8. [RFC: 2.6 patch] clarify the WIRELESS_EXT_SYSFS help text

    On Thu, Jul 24, 2008 at 01:38:40PM +0200, Johannes Berg wrote:
    > On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > > From commit 22bb1be4d27...:
    > >
    > > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > > by anyone."
    > >
    > > Can't have been much of an investigation as HAL uses these to determine
    > > if a card is a wireless one. Without this option, you cannot get
    > > wireless support in NetworkManager, stranding most laptops.

    >
    > Bzzzzt. You suck. Go check again. Hint: try hal's git tree.


    Can everyone agree on the patch below?

    > johannes


    cu
    Adrian


    <-- snip -->


    Current Hal uses the CONFIG_WIRELESS_EXT_SYSFS files, so don't claim
    there were no known users and recommend to enable the option.

    Signed-off-by: Adrian Bunk

    b685f62718b947264657c298e5ee5b2149200f68
    diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig
    index ab015c6..e582432 100644
    --- a/net/wireless/Kconfig
    +++ b/net/wireless/Kconfig
    @@ -39,4 +39,4 @@ config WIRELESS_EXT_SYSFS
    files in /sys/class/net/*/wireless/. The same information
    is available via the ioctls as well.

    - Say Y if you have programs using it (we don't know of any).
    + If unsure, say Y.
    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  9. Re: [RFC: 2.6 patch] clarify the WIRELESS_EXT_SYSFS help text

    On Thu, 2008-07-24 at 15:42 +0300, Adrian Bunk wrote:
    > On Thu, Jul 24, 2008 at 01:38:40PM +0200, Johannes Berg wrote:
    > > On Thu, 2008-07-24 at 11:58 +0200, Pierre Ossman wrote:
    > > > From commit 22bb1be4d27...:
    > > >
    > > > "The /sys/class/net/*/wireless/ direcory is, as far as I know, not used
    > > > by anyone."
    > > >
    > > > Can't have been much of an investigation as HAL uses these to determine
    > > > if a card is a wireless one. Without this option, you cannot get
    > > > wireless support in NetworkManager, stranding most laptops.

    > >
    > > Bzzzzt. You suck. Go check again. Hint: try hal's git tree.

    >
    > Can everyone agree on the patch below?
    >
    > > johannes

    >
    > cu
    > Adrian
    >
    >
    > <-- snip -->
    >
    >
    > Current Hal uses the CONFIG_WIRELESS_EXT_SYSFS files, so don't claim
    > there were no known users and recommend to enable the option.
    >
    > Signed-off-by: Adrian Bunk
    >
    > b685f62718b947264657c298e5ee5b2149200f68
    > diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig
    > index ab015c6..e582432 100644
    > --- a/net/wireless/Kconfig
    > +++ b/net/wireless/Kconfig
    > @@ -39,4 +39,4 @@ config WIRELESS_EXT_SYSFS
    > files in /sys/class/net/*/wireless/. The same information
    > is available via the ioctls as well.
    >
    > - Say Y if you have programs using it (we don't know of any).
    > + If unsure, say Y.


    Maybe also add:

    "HAL <= 0.5.11 depends on this feature."

    Where I hope I got the version number right :-)

    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  10. Re: [RFC: 2.6 patch] clarify the WIRELESS_EXT_SYSFS help text

    Sure, since we don't know what the next release will be there's little
    point in trying to predict it, this patch is fine with me.

    > Current Hal uses the CONFIG_WIRELESS_EXT_SYSFS files, so don't claim
    > there were no known users and recommend to enable the option.
    >
    > Signed-off-by: Adrian Bunk


    Acked-by: Johannes Berg

    >
    > b685f62718b947264657c298e5ee5b2149200f68
    > diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig
    > index ab015c6..e582432 100644
    > --- a/net/wireless/Kconfig
    > +++ b/net/wireless/Kconfig
    > @@ -39,4 +39,4 @@ config WIRELESS_EXT_SYSFS
    > files in /sys/class/net/*/wireless/. The same information
    > is available via the ioctls as well.
    >
    > - Say Y if you have programs using it (we don't know of any).
    > + If unsure, say Y.
    >


    -----BEGIN PGP SIGNATURE-----
    Comment: Johannes Berg (powerbook)

    iQIcBAABAgAGBQJIiORUAAoJEKVg1VMiehFYQvwQAJ8Y8VnaUP 4h4v3x304SAdJ3
    qbyPaPOCABE8ovadDoGYaOCslPdtY4P/vQ4FRlrmPU3Hu24F/cOB1JLpqDm249re
    /Foo2Bxh8nYeXqfqvC0NYpb7CtJyJwRAeN5ExUhfGYCO3THFPgQ VTmUnuI/TpJeg
    1552sIT31N5qnECR9Dj67V3R4qJJeTYG8y7UMvdu+4zj4xs5id w263BdSrmbFw9W
    bz+W51Q6j3qhn2cO+4dn22WWE5IltkcqbVEYcZbjreGYMStdJf G672fdEzKFioaU
    QAMKSuVYbkkTDcuO0gBMNvtm23+tC3mpVyAk+e8oYXFbBEqlj4 UUXXQlr+jiGPjQ
    ZocG9IKtNXYEpwKpSmjr068zhq8sh58G38LROQYyCksu9TNNRL cwThqijDueLeDK
    ZlaAWmp2jR0/edE//KFYlxvI45MDONGUQTDdiIbAYC4ql/n0zor4KoCZsCsf2sSn
    gP29/965rdfte07dlaGFsnU3tHjW1eofboyIdxhKPeFEnN7/y9bWvQ7Dymjz7Ybn
    z2IQB9jiuGb2z7dp1zMW5MD4cY8wRzXpB1eaJGdWYdik+Bevgs +PgR9SWSIb67u5
    aYnOXhuhrVJeLs81S1FHw5bx8rIm5YyKHE8TYybdtwLcUXGau4 QWo+6SNm6pAtz8
    iq0wzFEkZ4bHgVUNtwdk
    =J1rb
    -----END PGP SIGNATURE-----


+ Reply to Thread