Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH 1/2] pkg-cmake: add CMAKE_SYSTEM_VERSION to toolchainfile.cmake
Date: Sat, 12 Jan 2019 16:06:29 +0100	[thread overview]
Message-ID: <20190112160629.5a85748b@windsurf.home> (raw)
In-Reply-To: <20190111211626.GD2428@scaer>

Hello,

On Fri, 11 Jan 2019 22:16:26 +0100, Yann E. MORIN wrote:

> > the
> > value of BR2_TOOLCHAIN_HEADERS_AT_LEAST contains the (minimum) version
> > of the kernel it will run on,  
> 
> That is not true.
> 
> glibc for example will by default carry fallback mechanisms to emulate
> some features when it is running on a kernel older than the headers it
> was compiled against. In Buildroot we explictly disable that feature by
> passing --enable-kernel=$(BR2_TOOLCHAIN_HEADERS_AT_LEAST).
> 
> But for external toolchains, we don't know (and can't know) iw they did
> pass that option or not, and if they did, what version they requested,
> and if they did not, what version glibc uses.
> 
> This backward compatibility of glibc can extend quite a bit in the past,
> so kernels we would consider pretty old might still be usable with such
> toolchains. For example, on glibc-2.24, the minimum required headers
> were 3.2, but the runtime could be 2.6.32 on x86_64.
> 
> So, we could very well end up building a Buildroot system that runs on a
> kernel older than the headers of the toolchain.

While this is true, what can we do in practice ? We have no way to know
what will be the version of the kernel running on the target when we're
not building it, and still we have to provide this CMAKE_SYSTEM_VERSION
value. I am not really sure what packages are doing with
CMAKE_SYSTEM_VERSION. For example, azure-iot-sdk was simply testing
whether it was empty or not.

So for the time being, it seems like using the kernel headers version
is a good enough solution, even if not perfect.

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

  reply	other threads:[~2019-01-12 15:06 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-10 23:05 [Buildroot] [PATCH 1/2] pkg-cmake: add CMAKE_SYSTEM_VERSION to toolchainfile.cmake Arnout Vandecappelle
2019-01-10 23:05 ` [Buildroot] [PATCH 2/2] azure-iot-sdk-c: remove dummy value of CMAKE_SYSTEM_VERSION Arnout Vandecappelle
2019-01-11 21:16 ` [Buildroot] [PATCH 1/2] pkg-cmake: add CMAKE_SYSTEM_VERSION to toolchainfile.cmake Yann E. MORIN
2019-01-12 15:06   ` Thomas Petazzoni [this message]
2019-01-12 15:23     ` Yann E. MORIN
2019-01-12 16:07       ` Thomas Petazzoni
2019-01-12 15:07 ` Thomas Petazzoni
2019-01-12 15:26   ` Yann E. MORIN
2019-01-24 15:50   ` Peter Korsgaard

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=20190112160629.5a85748b@windsurf.home \
    --to=thomas.petazzoni@bootlin.com \
    --cc=buildroot@busybox.net \
    /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