Openembedded Devel Discussions
 help / color / mirror / Atom feed
From: Martin Jansa <martin.jansa@gmail.com>
To: openembedded-devel@lists.openembedded.org
Subject: Re: [PATCH] netmap-modules: make deterministic builds for drivers
Date: Wed, 18 Nov 2015 11:25:55 +0100	[thread overview]
Message-ID: <20151118102555.GB2213@jama> (raw)
In-Reply-To: <1447840941-29496-1-git-send-email-jackie.huang@windriver.com>

[-- Attachment #1: Type: text/plain, Size: 3710 bytes --]

On Wed, Nov 18, 2015 at 06:02:21PM +0800, jackie.huang@windriver.com wrote:
> From: Jackie Huang <jackie.huang@windriver.com>
> 
> The driver builds are optional, but for deterministic builds,
> we should should be able to explicitly enable/disable the
> builds for them in a proper place (maybe in BSP).
> But we can't use PACKAGECONFIG since there is no option for
> each driver, and the options are:
>  --no-drivers    do not compile any driver
>  --no-drivers=   do not compile the given drivers (comma sep.)
>  --drivers=      only compile the given drivers (comma sep.)
> 
> So use NETMAP_DRIVERS to list the needed drivers and add proper
> configs to EXTRA_OECONF, the default is no drivers, and all
> supported drivers are listed in NETMAP_ALL_DRIVERS.
> 
> Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
> ---
>  .../recipes-kernel/netmap/netmap-modules_git.bb    | 31 +++++++++++++++++++++-
>  1 file changed, 30 insertions(+), 1 deletion(-)
> 
> diff --git a/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb b/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
> index 6365fee..cfe5ce9 100644
> --- a/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
> +++ b/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
> @@ -15,7 +15,36 @@ EXTRA_OECONF = "--kernel-dir=${STAGING_KERNEL_BUILDDIR} \
>                  --driver-suffix="-netmap" \
>                  "
>  
> -EXTRA_OECONF += "--no-drivers=ixgbe --no-drivers=virtio_net.c"
> +# The driver builds are optional, but for deterministic builds,
> +# we should be able to explicitly enable/disable the builds
> +# for them in a proper place (maybe in BSP).
> +# But we can't use PACKAGECONFIG since there is no option for
> +# each driver, and the options are:
> +#  --no-drivers    do not compile any driver
> +#  --no-drivers=   do not compile the given drivers (comma sep.)
> +#  --drivers=      only compile the given drivers (comma sep.)
> +#
> +# So use NETMAP_DRIVERS and the following python code to add proper
> +# configs to EXTRA_OECONF and skip arch test for kernel modules.
> +#
> +# The default is no-drivers, and all supported drivers are listed
> +# in NETMAP_ALL_DRIVERS.
> +NETMAP_DRIVERS ??= ""
> +NETMAP_ALL_DRIVERS = "ixgbe igb e1000e e1000 veth.c forcedeth.c virtio_net.c r8169.c"
> +
> +python __anonymous () {
> +    drivers_list = d.getVar("NETMAP_DRIVERS", True).split()
> +    all_drivers_list = d.getVar("NETMAP_ALL_DRIVERS", True).split()
> +    config_drivers = "--drivers=" + ",".join(drivers_list)
> +
> +    extra_oeconf_drivers = bb.utils.contains_any('NETMAP_DRIVERS', all_drivers_list, config_drivers, '--no-drivers', d)
> +    d.appendVar("EXTRA_OECONF", extra_oeconf_drivers)

This part looks good to me, thanks for fixing virtio_net.c change (I've
merged that one by accident).

> +
> +    # skip the arch test for kernel modules
> +    if drivers_list:
> +        for driver in drivers_list:
> +            d.setVar("INSANE_SKIP_kernel-module-%s-netmap" % driver, "arch")

Why is this needed? Shouldn't the kernel modules build with correct
architecture? And e.g. r8169.c creates kernel-module-r8169.c-netmap package?
I hope the .c is stripped somewhere in the build.


> +}
>  
>  LDFLAGS := "${@'${LDFLAGS}'.replace('-Wl,-O1', '')}"
>  LDFLAGS := "${@'${LDFLAGS}'.replace('-Wl,--as-needed', '')}"
> -- 
> 1.9.1
> 
> -- 
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-devel

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 188 bytes --]

  reply	other threads:[~2015-11-18 10:25 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-18 10:02 [PATCH] netmap-modules: make deterministic builds for drivers jackie.huang
2015-11-18 10:25 ` Martin Jansa [this message]
2015-11-19  3:20   ` Huang, Jie (Jackie)

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20151118102555.GB2213@jama \
    --to=martin.jansa@gmail.com \
    --cc=openembedded-devel@lists.openembedded.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox