Patchwork bitbake.conf/abi_version.conf: Update to use BasicHash signature generator

login
register
mail settings
Submitter Richard Purdie
Date June 22, 2012, 2:06 p.m.
Message ID <1340373998.394.36.camel@ted>
Download mbox | patch
Permalink /patch/30471/
State Accepted
Commit 4199efed48005a62267fa3374c33b13627d85f44
Headers show

Comments

Richard Purdie - June 22, 2012, 2:06 p.m.
This switches to use the hash based signature generatior and update
layout_abi to match stamp file layout changes.

Angstrom, SHR and Poky did this a while ago. This brings the OE-Core defaults
into line with what is the best common practise at this point. Its been
discussed on the mailing lists and by the TSC at length, this just completes
the transition.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
---
 meta/conf/abi_version.conf |    2 +-
 meta/conf/bitbake.conf     |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
Khem Raj - June 22, 2012, 2:18 p.m.
On Fri, Jun 22, 2012 at 7:06 AM, Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:
> This switches to use the hash based signature generatior and update
> layout_abi to match stamp file layout changes.
>
> Angstrom, SHR and Poky did this a while ago. This brings the OE-Core defaults
> into line with what is the best common practise at this point. Its been
> discussed on the mailing lists and by the TSC at length, this just completes
> the transition.
>
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

thanks looks good


> ---
>  meta/conf/abi_version.conf |    2 +-
>  meta/conf/bitbake.conf     |    2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/meta/conf/abi_version.conf b/meta/conf/abi_version.conf
> index f2919e5..83b668a 100644
> --- a/meta/conf/abi_version.conf
> +++ b/meta/conf/abi_version.conf
> @@ -4,4 +4,4 @@
>  # that breaks the format and have been previously discussed on the mailing list
>  # with general agreement from the core team.
>  #
> -OELAYOUT_ABI = "7"
> +OELAYOUT_ABI = "8"
> diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
> index b6fbc7a..fbc0ca2 100644
> --- a/meta/conf/bitbake.conf
> +++ b/meta/conf/bitbake.conf
> @@ -748,7 +748,7 @@ TRANSLATED_TARGET_ARCH ??= "${@d.getVar('TARGET_ARCH', True).replace("_", "-")}"
>  BB_CONSOLELOG ?= "${TMPDIR}/cooker.log.${DATETIME}"
>
>  # Setup our default hash policy
> -BB_SIGNATURE_HANDLER ?= "OEBasic"
> +BB_SIGNATURE_HANDLER ?= "OEBasicHash"
>  BB_HASHBASE_WHITELIST ?= "TMPDIR FILE PATH PWD BB_TASKHASH BBPATH DL_DIR SSTATE_DIR THISDIR FILESEXTRAPATHS FILE_DIRNAME HOME LOGNAME SHELL TERM USER FILESPATH STAGING_DIR_HOST STAGING_DIR_TARGET COREBASE PRSERV_HOST PRSERV_PORT PRSERV_DUMPDIR PRSERV_DUMPFILE PRSERV_LOCKDOWN PARALLEL_MAKE CCACHE_DIR EXTERNAL_TOOLCHAIN CCACHE"
>  BB_HASHCONFIG_WHITELIST ?= "${BB_HASHBASE_WHITELIST} DATE TIME SESSION_MANAGER DBUS_SESSION_BUS_ADDRESS SSH_AGENT_PID XDG_SESSION_COOKIE SSH_AUTH_SOCK XAUTHORITY PSEUDO_BUILD"
>  BB_SIGNATURE_EXCLUDE_FLAGS ?= "doc defaultval _append _prepend deps depends lockfiles type vardepsexclude \
>
>
>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Enrico Scholz - June 22, 2012, 2:24 p.m.
Richard Purdie
<richard.purdie-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r@public.gmane.org>
writes:

> This switches to use the hash based signature generatior and update
> layout_abi to match stamp file layout changes.

Has this really been tested? I played a little bit with it and it was
very painful when whole distribution will be rebuilt due to a minor
change in an early package.

There are some subtle bugs in the hash generation too (e.g. bitbake's
get_file_depends() result is used to calculate a hash over a set() which
is filled nondeterministically when BB_NUMBER_THREADS>1) which cause
random rebuilds too.



Enrico
Khem Raj - June 22, 2012, 2:26 p.m.
On Fri, Jun 22, 2012 at 7:24 AM, Enrico Scholz
<enrico.scholz@sigma-chemnitz.de> wrote:
> Has this really been tested?

Angstrom. SHR, yocto has it on by default for months now. No one
reported issues thus far.
Martin Jansa - June 22, 2012, 2:31 p.m.
On Fri, Jun 22, 2012 at 03:06:38PM +0100, Richard Purdie wrote:
> This switches to use the hash based signature generatior and update
> layout_abi to match stamp file layout changes.
> 
> Angstrom, SHR and Poky did this a while ago. This brings the OE-Core defaults

SHR didn't (except my testing), but I'm fine with this change.. I'll override it 
in back to OEBasic in meta-shr layer (at least until we get faster
official feed builder..).

Cheers,

> into line with what is the best common practise at this point. Its been
> discussed on the mailing lists and by the TSC at length, this just completes
> the transition.
> 
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> ---
>  meta/conf/abi_version.conf |    2 +-
>  meta/conf/bitbake.conf     |    2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/meta/conf/abi_version.conf b/meta/conf/abi_version.conf
> index f2919e5..83b668a 100644
> --- a/meta/conf/abi_version.conf
> +++ b/meta/conf/abi_version.conf
> @@ -4,4 +4,4 @@
>  # that breaks the format and have been previously discussed on the mailing list 
>  # with general agreement from the core team.
>  #
> -OELAYOUT_ABI = "7"
> +OELAYOUT_ABI = "8"
> diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
> index b6fbc7a..fbc0ca2 100644
> --- a/meta/conf/bitbake.conf
> +++ b/meta/conf/bitbake.conf
> @@ -748,7 +748,7 @@ TRANSLATED_TARGET_ARCH ??= "${@d.getVar('TARGET_ARCH', True).replace("_", "-")}"
>  BB_CONSOLELOG ?= "${TMPDIR}/cooker.log.${DATETIME}"
>  
>  # Setup our default hash policy
> -BB_SIGNATURE_HANDLER ?= "OEBasic"
> +BB_SIGNATURE_HANDLER ?= "OEBasicHash"
>  BB_HASHBASE_WHITELIST ?= "TMPDIR FILE PATH PWD BB_TASKHASH BBPATH DL_DIR SSTATE_DIR THISDIR FILESEXTRAPATHS FILE_DIRNAME HOME LOGNAME SHELL TERM USER FILESPATH STAGING_DIR_HOST STAGING_DIR_TARGET COREBASE PRSERV_HOST PRSERV_PORT PRSERV_DUMPDIR PRSERV_DUMPFILE PRSERV_LOCKDOWN PARALLEL_MAKE CCACHE_DIR EXTERNAL_TOOLCHAIN CCACHE"
>  BB_HASHCONFIG_WHITELIST ?= "${BB_HASHBASE_WHITELIST} DATE TIME SESSION_MANAGER DBUS_SESSION_BUS_ADDRESS SSH_AGENT_PID XDG_SESSION_COOKIE SSH_AUTH_SOCK XAUTHORITY PSEUDO_BUILD"
>  BB_SIGNATURE_EXCLUDE_FLAGS ?= "doc defaultval _append _prepend deps depends lockfiles type vardepsexclude \
> 
> 
> 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Richard Purdie - June 22, 2012, 2:40 p.m.
On Fri, 2012-06-22 at 16:31 +0200, Martin Jansa wrote:
> On Fri, Jun 22, 2012 at 03:06:38PM +0100, Richard Purdie wrote:
> > This switches to use the hash based signature generatior and update
> > layout_abi to match stamp file layout changes.
> > 
> > Angstrom, SHR and Poky did this a while ago. This brings the OE-Core defaults
> 
> SHR didn't (except my testing), but I'm fine with this change.. I'll override it 
> in back to OEBasic in meta-shr layer (at least until we get faster
> official feed builder..).

Sorry, I thought you had switched, my mistake. I'll fix that commit
message before merging.

Cheers,

Richard
Richard Purdie - June 22, 2012, 2:42 p.m.
On Fri, 2012-06-22 at 16:24 +0200, Enrico Scholz wrote:
> Richard Purdie
> <richard.purdie-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r@public.gmane.org>
> writes:
> 
> > This switches to use the hash based signature generatior and update
> > layout_abi to match stamp file layout changes.
> 
> Has this really been tested?

Its the default in Poky and Angstrom. It was the default for the Yocto
Project 1.2 release too. So yes, it has seen testing.

>  I played a little bit with it and it was
> very painful when whole distribution will be rebuilt due to a minor
> change in an early package.
> 
> There are some subtle bugs in the hash generation too (e.g. bitbake's
> get_file_depends() result is used to calculate a hash over a set() which
> is filled nondeterministically when BB_NUMBER_THREADS>1) which cause
> random rebuilds too.

Bug reports on issues like this would be good, we can't fix a problem if
we don't know about it.

Cheers,

Richard
Enrico Scholz - June 22, 2012, 3:11 p.m.
Richard Purdie
<richard.purdie-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r@public.gmane.org>
writes:

>> There are some subtle bugs in the hash generation too (e.g. bitbake's
>> get_file_depends() result is used to calculate a hash over a set() which
>> is filled nondeterministically when BB_NUMBER_THREADS>1) which cause
>> random rebuilds too.
>
> Bug reports on issues like this would be good, we can't fix a problem if
> we don't know about it.

http://lists.linuxtogo.org/pipermail/bitbake-devel/2012-May/002951.html


Enrico

Patch

diff --git a/meta/conf/abi_version.conf b/meta/conf/abi_version.conf
index f2919e5..83b668a 100644
--- a/meta/conf/abi_version.conf
+++ b/meta/conf/abi_version.conf
@@ -4,4 +4,4 @@ 
 # that breaks the format and have been previously discussed on the mailing list 
 # with general agreement from the core team.
 #
-OELAYOUT_ABI = "7"
+OELAYOUT_ABI = "8"
diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
index b6fbc7a..fbc0ca2 100644
--- a/meta/conf/bitbake.conf
+++ b/meta/conf/bitbake.conf
@@ -748,7 +748,7 @@  TRANSLATED_TARGET_ARCH ??= "${@d.getVar('TARGET_ARCH', True).replace("_", "-")}"
 BB_CONSOLELOG ?= "${TMPDIR}/cooker.log.${DATETIME}"
 
 # Setup our default hash policy
-BB_SIGNATURE_HANDLER ?= "OEBasic"
+BB_SIGNATURE_HANDLER ?= "OEBasicHash"
 BB_HASHBASE_WHITELIST ?= "TMPDIR FILE PATH PWD BB_TASKHASH BBPATH DL_DIR SSTATE_DIR THISDIR FILESEXTRAPATHS FILE_DIRNAME HOME LOGNAME SHELL TERM USER FILESPATH STAGING_DIR_HOST STAGING_DIR_TARGET COREBASE PRSERV_HOST PRSERV_PORT PRSERV_DUMPDIR PRSERV_DUMPFILE PRSERV_LOCKDOWN PARALLEL_MAKE CCACHE_DIR EXTERNAL_TOOLCHAIN CCACHE"
 BB_HASHCONFIG_WHITELIST ?= "${BB_HASHBASE_WHITELIST} DATE TIME SESSION_MANAGER DBUS_SESSION_BUS_ADDRESS SSH_AGENT_PID XDG_SESSION_COOKIE SSH_AUTH_SOCK XAUTHORITY PSEUDO_BUILD"
 BB_SIGNATURE_EXCLUDE_FLAGS ?= "doc defaultval _append _prepend deps depends lockfiles type vardepsexclude \