Re: [gentoo-dev] [PATCH] savedconfig.eclass: clean up ED and EROOT usage

2019-05-23 Thread Mike Gilbert
On Fri, May 24, 2019 at 1:24 AM Ulrich Mueller  wrote:
>
> > On Fri, 24 May 2019, Mike Gilbert wrote:
>
> > On Thu, May 23, 2019 at 7:16 PM David Seifert  wrote:
> >> Given that there are no ebuilds in the tree using this eclass and being
> >> in EAPI 0, 1 or 2 (
> >> https://qa-reports.gentoo.org/output/eapi-per-eclass/savedconfig.eclass/
> >> ), wouldn't it make more sense to just whitelist EAPI >= 4 and clean up
> >> this backwards compatibility cruft instead?
>
> > I'm fixing a bug with the least invasive change possible. I'm not
> > trying to rework the eclass.
>
> AFAICS, that backwards compatibility code consists of two case
> statements, and the chance that removing them would break anything is
> close to zero. So I wouldn't call it a "rework". :)
>
> I'd rather remove than update that code for deprecated EAPIs. No ebuild
> would ever use it, so your updated code would never be tested.

Again, I'm fixing a bug. Removing EAPI 0-2 compatibility is
unnecessary to fix the bug.



Re: [gentoo-dev] [PATCH] savedconfig.eclass: clean up ED and EROOT usage

2019-05-23 Thread Ulrich Mueller
> On Fri, 24 May 2019, Mike Gilbert wrote:

> On Thu, May 23, 2019 at 7:16 PM David Seifert  wrote:
>> Given that there are no ebuilds in the tree using this eclass and being
>> in EAPI 0, 1 or 2 (
>> https://qa-reports.gentoo.org/output/eapi-per-eclass/savedconfig.eclass/
>> ), wouldn't it make more sense to just whitelist EAPI >= 4 and clean up
>> this backwards compatibility cruft instead?

> I'm fixing a bug with the least invasive change possible. I'm not
> trying to rework the eclass.

AFAICS, that backwards compatibility code consists of two case
statements, and the chance that removing them would break anything is
close to zero. So I wouldn't call it a "rework". :)

I'd rather remove than update that code for deprecated EAPIs. No ebuild
would ever use it, so your updated code would never be tested.

Ulrich


signature.asc
Description: PGP signature


Re: [gentoo-dev] [PATCH 0/4] llvm.eclass: EAPI 7 support

2019-05-23 Thread Michał Górny
On Tue, 2019-04-30 at 07:38 +0200, Michał Górny wrote:
> Hi,
> 
> Here's my proposed EAPI 7 support for llvm.eclass.  I think it should
> conceptually work with cross but I haven't tested it.  The main problem
> is that we have multiple distinct ways of building LLVM, and AFAIU
> when using CMake modules, you should be able to get away without having
> LLVM in CBUILD.  However, when using llvm-config you obviously need
> it CBUILD-executable.
> 
> To support both scenarios, get_llvm_prefix is now provided with '-b'
> and '-d' options (matching those to has_version).  Using them, you can
> choose the behavior suitable for your package.
> 
> I've left the default at '-d'.  I don't really know which is better
> for the generic cross case but either way should work for non-cross
> setups.
> 
> --
> Best regards,
> Michał Górny
> 
> Michał Górny (4):
>   llvm.eclass: Remove unnecessary '_rc' from < example
>   llvm.eclass: Update examples for newer LLVM versions
>   llvm.eclass: Add EAPI 7 API to get_llvm_prefix
>   llvm.eclass: Enable EAPI 7
> 
>  eclass/llvm.eclass | 65 +-
>  1 file changed, 53 insertions(+), 12 deletions(-)
> 

Merged.

-- 
Best regards,
Michał Górny



signature.asc
Description: This is a digitally signed message part


Re: [gentoo-dev] toolchain.eclass musl changes

2019-05-23 Thread Jory A. Pratt
On 5/21/19 3:08 AM, Sergei Trofimovich wrote:
> On Mon, 20 May 2019 16:41:56 -0500
> "Jory A. Pratt"  wrote:
> 
>> Please review the proposed changes for musl support. We would like to
>> get this landed on the tree ASAP so we can drop the eclass from the musl
>> overlay. This is one of the first steps to getting things moved to tree
>> so we can depreciate the overlay.
> 
> Looks ok.
> 
Thank you, I have committed the change to tree.

Jory



signature.asc
Description: OpenPGP digital signature


[gentoo-dev] Last rites: net-misc/mindterm

2019-05-23 Thread Aaron Bauman
# Aaron Bauman  (23 May 2019)
# Upstreami now requires registration to download
# EAPI2. Removal in 30 days.
net-misc/mindterm

-- 
Cheers,
Aaron


signature.asc
Description: PGP signature


Re: [gentoo-dev] Last rites: dev-util/pmd

2019-05-23 Thread Aaron Bauman
On Thu, May 23, 2019 at 08:58:15PM -0400, Aaron Bauman wrote:
> # Aaron Bauman  (23 May 2019)
> # Unmaintained, EAPI2, updates available upstream
> dev-util/pmd
> 
> -- 
> Cheers,
> Aaron

Oh yea, removal in 30 days.

-- 
Cheers,
Aaron


signature.asc
Description: PGP signature


Re: [gentoo-dev] [PATCH] savedconfig.eclass: clean up ED and EROOT usage

2019-05-23 Thread Mike Gilbert
On Thu, May 23, 2019 at 7:16 PM David Seifert  wrote:
> Given that there are no ebuilds in the tree using this eclass and being
> in EAPI 0, 1 or 2 (
> https://qa-reports.gentoo.org/output/eapi-per-eclass/savedconfig.eclass/
> ), wouldn't it make more sense to just whitelist EAPI >= 4 and clean up
> this backwards compatibility cruft instead?

I'm fixing a bug with the least invasive change possible. I'm not
trying to rework the eclass.



[gentoo-dev] Last rites: dev-util/pmd

2019-05-23 Thread Aaron Bauman
# Aaron Bauman  (23 May 2019)
# Unmaintained, EAPI2, updates available upstream
dev-util/pmd

-- 
Cheers,
Aaron


signature.asc
Description: PGP signature


Re: [gentoo-dev] [PATCH] savedconfig.eclass: clean up ED and EROOT usage

2019-05-23 Thread David Seifert
On Thu, 2019-05-23 at 11:15 -0400, Mike Gilbert wrote:
> Avoid assigning these variables in EAPIs where they are already
> defined.
> Also make them local variables to avoid polluting the global
> environment.
> 
> Closes: https://bugs.gentoo.org/685382
> ---
>  eclass/savedconfig.eclass | 18 ++
>  1 file changed, 10 insertions(+), 8 deletions(-)
> 
> diff --git a/eclass/savedconfig.eclass b/eclass/savedconfig.eclass
> index 1e9ac6c80b4d..9bd308685b2d 100644
> --- a/eclass/savedconfig.eclass
> +++ b/eclass/savedconfig.eclass
> @@ -46,19 +46,20 @@ save_config() {
>   fi
>   [[ $# -eq 0 ]] && die "Usage: save_config "
>  
> - # Be lazy in our EAPI compat
> - : ${ED:=${D}}
> + case ${EAPI:-0} in
> + 0|1|2) local ED=${D}
> + esac
>  
>   local dest="/etc/portage/savedconfig/${CATEGORY}"
>   if [[ $# -eq 1 && -f $1 ]] ; then
>   # Just one file, so have the ${PF} be that config file
>   dodir "${dest}"
> - cp "$@" "${ED}/${dest}/${PF}" || die "failed to save
> $*"
> + cp "$@" "${ED%/}/${dest}/${PF}" || die "failed to save
> $*"
>   else
>   # A dir, or multiple files, so have the ${PF} be a dir
>   # with all the saved stuff below it
>   dodir "${dest}/${PF}"
> - treecopy "$@" "${ED}/${dest}/${PF}" || die "failed to
> save $*"
> + treecopy "$@" "${ED%/}/${dest}/${PF}" || die "failed to
> save $*"
>   fi
>  
>   elog "Your configuration for ${CATEGORY}/${PF} has been saved
> in "
> @@ -99,7 +100,7 @@ restore_config() {
>   use savedconfig || return
>  
>   local found check configfile
> - local base=${PORTAGE_CONFIGROOT}/etc/portage/savedconfig
> + local base=${PORTAGE_CONFIGROOT%/}/etc/portage/savedconfig
>   for check in
> {${CATEGORY}/${PF},${CATEGORY}/${P},${CATEGORY}/${PN}}; do
>   configfile=${base}/${CTARGET}/${check}
>   [[ -r ${configfile} ]] ||
> configfile=${base}/${CHOST}/${check}
> @@ -143,10 +144,11 @@ savedconfig_pkg_postinst() {
>   # are worse :/.
>  
>   if use savedconfig ; then
> - # Be lazy in our EAPI compat
> - : ${EROOT:=${ROOT}}
> + case ${EAPI:-0} in
> + 0|1|2) local EROOT=${ROOT}
> + esac
>  
> - find
> "${EROOT}/etc/portage/savedconfig/${CATEGORY}/${PF}" \
> + find
> "${EROOT%/}/etc/portage/savedconfig/${CATEGORY}/${PF}" \
>   -exec touch {} + 2>/dev/null
>   fi
>  }

Given that there are no ebuilds in the tree using this eclass and being
in EAPI 0, 1 or 2 (
https://qa-reports.gentoo.org/output/eapi-per-eclass/savedconfig.eclass/
), wouldn't it make more sense to just whitelist EAPI >= 4 and clean up
this backwards compatibility cruft instead?




Re: [gentoo-dev] Packages up for grabs due to nimiux's retirement

2019-05-23 Thread Patrice Clement
Monday 20 May 2019 17:53:52, Michał Górny wrote :
> Hello,
> 
> The following packages are now up for grabs:
> 
> [...]
> app-misc/vifm
>
I'm taking over this one.

-- 
Patrice Clement
Gentoo Linux developer
http://www.gentoo.org


signature.asc
Description: PGP signature


Re: [gentoo-dev] Packages up for grabs due to nimiux's retirement

2019-05-23 Thread Bernard Cafarelli
Le Mon, 20 May 2019 17:53:52 +0200
Michał Górny  a écrit:

> Hello,
> 
> The following packages are now up for grabs:
> 
> app-portage/conf-update
> app-admin/logrotate
> app-admin/mktwpol
> app-admin/swatchdog
> app-admin/tripwire
> app-emulation/free42
I will take this HP calc...
> app-emulation/libdsk
> app-emulation/x48
... and this one too
> app-emulation/xcpc
> app-misc/linux-logo
Nostalgia striking I will keep this one around
> app-misc/muttprint
> app-misc/vifm
> x11-wm/stumpwm-contrib
> x11-wm/stumpwm
> 
> -- 
> Best regards,
> Michał Górny
> 


-- 
Bernard Cafarelli (Voyageur)
Gentoo developer



[gentoo-dev] [PATCH] savedconfig.eclass: clean up ED and EROOT usage

2019-05-23 Thread Mike Gilbert
Avoid assigning these variables in EAPIs where they are already defined.
Also make them local variables to avoid polluting the global
environment.

Closes: https://bugs.gentoo.org/685382
---
 eclass/savedconfig.eclass | 18 ++
 1 file changed, 10 insertions(+), 8 deletions(-)

diff --git a/eclass/savedconfig.eclass b/eclass/savedconfig.eclass
index 1e9ac6c80b4d..9bd308685b2d 100644
--- a/eclass/savedconfig.eclass
+++ b/eclass/savedconfig.eclass
@@ -46,19 +46,20 @@ save_config() {
fi
[[ $# -eq 0 ]] && die "Usage: save_config "
 
-   # Be lazy in our EAPI compat
-   : ${ED:=${D}}
+   case ${EAPI:-0} in
+   0|1|2) local ED=${D}
+   esac
 
local dest="/etc/portage/savedconfig/${CATEGORY}"
if [[ $# -eq 1 && -f $1 ]] ; then
# Just one file, so have the ${PF} be that config file
dodir "${dest}"
-   cp "$@" "${ED}/${dest}/${PF}" || die "failed to save $*"
+   cp "$@" "${ED%/}/${dest}/${PF}" || die "failed to save $*"
else
# A dir, or multiple files, so have the ${PF} be a dir
# with all the saved stuff below it
dodir "${dest}/${PF}"
-   treecopy "$@" "${ED}/${dest}/${PF}" || die "failed to save $*"
+   treecopy "$@" "${ED%/}/${dest}/${PF}" || die "failed to save $*"
fi
 
elog "Your configuration for ${CATEGORY}/${PF} has been saved in "
@@ -99,7 +100,7 @@ restore_config() {
use savedconfig || return
 
local found check configfile
-   local base=${PORTAGE_CONFIGROOT}/etc/portage/savedconfig
+   local base=${PORTAGE_CONFIGROOT%/}/etc/portage/savedconfig
for check in {${CATEGORY}/${PF},${CATEGORY}/${P},${CATEGORY}/${PN}}; do
configfile=${base}/${CTARGET}/${check}
[[ -r ${configfile} ]] || configfile=${base}/${CHOST}/${check}
@@ -143,10 +144,11 @@ savedconfig_pkg_postinst() {
# are worse :/.
 
if use savedconfig ; then
-   # Be lazy in our EAPI compat
-   : ${EROOT:=${ROOT}}
+   case ${EAPI:-0} in
+   0|1|2) local EROOT=${ROOT}
+   esac
 
-   find "${EROOT}/etc/portage/savedconfig/${CATEGORY}/${PF}" \
+   find "${EROOT%/}/etc/portage/savedconfig/${CATEGORY}/${PF}" \
-exec touch {} + 2>/dev/null
fi
 }
-- 
2.21.0