Re: [gentoo-user] GCC 5.4.0

[Dale]

Dale wrote:
>
> I just did this myself.  After I switched to the new gcc, I ran
> revdep-rebuild.  It had a large list of packages.  I was planning to do
> a emerge -e world anyway, so I just did it instead.  No sense doing most
> of it twice.  During the rebuild, I had a qt package to fail.  For some
> reason, it was stuck on a old qt4 version, even tho a qt5 version was
> there and ready to upgrade.  I went ahead and manually upgraded it then
> restarted the emerge process again, since it didn't build many packages
> before that one failed anyway.  It completed the whole emerge without a
> single failure that time. 
>
> I will say this, I had to switch to Fluxbox for a while.  KDE got pretty
> weird.  I logged out, back in and it was really weird.  I got a few
> pop-ups about things not starting correctly etc.  It was unusable at
> that point.  I then went to Fluxbox and used it for a while.  Once the
> emerge got mostly done, KDE started working correctly again.  As I
> suspected, some things just didn't like the difference in the gcc
> versions.  No surprise there really. 
>
> The only hitch so far, digikam is complaining with this error. 
>
> digikam: symbol lookup error: /usr/lib64/libdigikamcore.so.5.5.0:
> undefined symbol:
>
> I'm currently rebuilding it after revdep-rebuild said it needed it.  I
> suspect it will work once that is done.  I hope so.  I got pics to
> download from my camera. 
>
> All in all, it got weird for a bit but in the end, it went fairly well. 
> The one failure I had wasn't related to the gcc upgrade.  I still have
> no idea why that one package was stuck on that qt4 version.  For anyone
> using KDE and doing this, I'd do it from a console and have a time where
> you either have a backup desktop to use or some time to let it sit and
> compile and at least get most of KDE re-emerged.  For me at least, KDE
> was unusable for a while there.
>
> It's amazing that emerge can compile about 1500 packages and not have a
> failure.  Those devs are really doing some good work back there.  :-D 
>
> Oh, I skipped palemoon.  I'm not using it and may unmerge it anyway.  I
> almost forgot that one.  That is talked about elsewhere in this thread. 
>
> Dale
>
> :-)  :-) 
>


I wanted to add one more thing to this and see if anyone else noticed
the same thing.  I've mentioned before that I use several Firefox
profiles.  In other words, I have several instances of Firefox running
at the same time doing different things.  A couple or so of those can
have over 100 tabs open at one time. 

This is something I have noticed that changed since the gcc upgrade. 
One, Firefox seems to use less memory.  It's not a whole lot less but it
is less for sure.  It also uses a lot less CPU power.  Used to if I had
two or three Firefox profiles running, I could see 20, 30 and sometimes
even more CPU usage.  Gkrellm would have a good bit of orange and the
little needle thing would be hovering around 30 to 40% on all the cores
with it spiking to well over 50% quite a lot.  Closing Firefox would put
it down to single digits so I know it was Firefox causing this.  Now,
with 4 instances running, it is hovering around 10 to 15%. 

Second, I notice the same with plasmashell as well.  It used to get
pretty hungry on memory and CPU at times.  After the upgrade, not to
bad.  Seems to stay about the same even after not logging out for days
at a time. 

I did do a recent Firefox upgrade and a KDE upgrade as well.  I'm not
100% sure what has the most effect on this but the two programs aren't
related in any way except that they were built with the newer gcc.  I
found it interesting and was curious if others who upgraded have noticed
anything similar to this.  Maybe noticed it with other programs as well. 

Dale

:-)  :-)

Re: [gentoo-user] ebuild: package specific CFLAGS

[John Covici]

On Fri, 28 Apr 2017 22:10:42 -0400,
Ian Zimmerman wrote:
> 
> I'm trying to create an ebuild of a crufty old program that needs
> -fgnu89-inline in compiler flags to have any chance of building.
> 
> What's the way to do that in an ebuild?  I could have something like
> 
> src_configure() {
> econf $(use_enable nls) CFLAGS=-fgnu89-inline
> }
> 
> but then, will this not _override_ (rather than add to, as desired) the
> CFLAGS from make.conf?

Maybe you'd be better off setting an environment variable outside the
ebuild in a shell script in /etc/portage/env where you can put the
whole CCFLAGS .

-- 
Your life is like a penny.  You're going to lose it.  The question is:
How do
you spend it?

 John Covici
 cov...@ccs.covici.com

Re: [gentoo-user] Re: In search of an truecolor-capable terminal emulator

[tuxic]

On 04/28 07:32, Fernando Rodriguez wrote:
> On 04/28/2017 02:59 PM, Ian Zimmerman wrote:
> > On 2017-04-28 10:10, Fernando Rodriguez wrote:
> > 
> > > No. I meant you can't enable them *all* globally, meaning opengl,
> > > gles, egl, etc. It's kind of the same situation as with GUI toolkits,
> > > you can't enable them all globally because some packages support more
> > > than one that you have to choose at compile time.
> > 
> > I think we need to spell out what you mean by this:
> > 
> > > > > if any of them have egl or gles and not opengl then enable it.
> > 
> > by "not having opengl" do you mean that it's not in the package's IUSE
> > at all, or that it is disabled?
> 
> On the sentence prior to the one you quoted I advised the OP to enable
> opengl globally. Therefore, if a package has it then it is already enabled.
> So obviously I meant if it doesn't have it at all.
> 
> > by "enable *it*" do you mean enable opengl, or enable those other flags?
> 
> For the same reason stated above I can only be refering to those other
> flags. Those are the flags that control GL acceleration. If you want your
> system to use your GPU as much as possible then you need to enable at least
> one of them on all the packages that support it. If a package has more than
> one it's between you and portage which one you choose because some packages
> depend on one or the other. So I find it easier to enable opengl first and
> then work through the conflicts to enable the others as much as possible.
> 
> -- 
> 
> Fernando Rodriguez
> 

Hi,

before changing my system I took a look at
https://wiki.gentoo.org/wiki/Xorg/Hardware_3D_acceleration_guide

Down the page there is a test, whohc on my system reports:
glxinfo | grep rendering
direct rendering: Yes
GL_NV_path_rendering, GL_NV_pixel_data_range, GL_NV_point_sprite, 
GL_NV_path_rendering, GL_NV_pixel_data_range, GL_NV_point_sprite, 
GL_NV_packed_float_linear, GL_NV_path_rendering, 


For me it reads like: Yes, you have hardware accelerated rendering.

Or did I miss something...?

Cheers
Meino

[gentoo-user] ebuild: package specific CFLAGS

[Ian Zimmerman]

I'm trying to create an ebuild of a crufty old program that needs
-fgnu89-inline in compiler flags to have any chance of building.

What's the way to do that in an ebuild?  I could have something like

src_configure() {
econf $(use_enable nls) CFLAGS=-fgnu89-inline
}

but then, will this not _override_ (rather than add to, as desired) the
CFLAGS from make.conf?

-- 
Please *no* private Cc: on mailing lists and newsgroups
Personal signed mail: please _encrypt_ and sign
Don't clear-text sign:
http://primate.net/~itz/blog/the-problem-with-gpg-signatures.html

Re: [gentoo-user] New AMD hardware. Still can't boot (but making progress).

[Bill Kenworthy]

>>> -- 
>>> Regards,
>>> Mick
> 
> Have you tried connecting using ssh after boot?
> Also, do you have the EFI console support in your kernel?
> 
> Not near my desktop to check which ones exactly.
> 
> --
> Joost
> 

Second this - check the kernel options for EFI but also add the NVME drives.

I failed to get grub working (rather, gave up because of the lack of
docs) and went with rEFIt.

BillK

MS Surface Pro 4

bunyip ~ # grep NVME /usr/src/linux/.config
CONFIG_NVME_CORE=y
CONFIG_BLK_DEV_NVME=y
CONFIG_BLK_DEV_NVME_SCSI=y
CONFIG_NVME_FABRICS=y
# CONFIG_NVME_FC is not set
CONFIG_NVME_TARGET=y
CONFIG_NVME_TARGET_LOOP=y
# CONFIG_NVME_TARGET_FC is not set
CONFIG_NVMEM=m
bunyip ~ #


bunyip ~ # grep EFI /usr/src/linux/.config
CONFIG_EFI_PARTITION=y
CONFIG_EFI=y
CONFIG_EFI_STUB=y
CONFIG_EFI_MIXED=y
CONFIG_FB_EFI=y
CONFIG_DMI_SCAN_MACHINE_NON_EFI_FALLBACK=y
# EFI (Extensible Firmware Interface) Support
CONFIG_EFI_VARS=y
CONFIG_EFI_ESRT=y
CONFIG_EFI_FAKE_MEMMAP=y
CONFIG_EFI_MAX_FAKE_MEM=8
CONFIG_EFI_RUNTIME_WRAPPERS=y
CONFIG_EFI_BOOTLOADER_CONTROL=y
# CONFIG_EFI_CAPSULE_LOADER is not set
# CONFIG_EFI_TEST is not set
# CONFIG_EFI_DEV_PATH_PARSER is not set
CONFIG_CACHEFILES=m
# CONFIG_CACHEFILES_DEBUG is not set
# CONFIG_CACHEFILES_HISTOGRAM is not set
CONFIG_EFIVAR_FS=y
CONFIG_EARLY_PRINTK_EFI=y
# CONFIG_EFI_PGT_DUMP is not set
bunyip ~ #

Re: [gentoo-user] Re: In search of an truecolor-capable terminal emulator

[Fernando Rodriguez]

On 04/28/2017 02:59 PM, Ian Zimmerman wrote:

On 2017-04-28 10:10, Fernando Rodriguez wrote:


No. I meant you can't enable them *all* globally, meaning opengl,
gles, egl, etc. It's kind of the same situation as with GUI toolkits,
you can't enable them all globally because some packages support more
than one that you have to choose at compile time.


I think we need to spell out what you mean by this:


if any of them have egl or gles and not opengl then enable it.


by "not having opengl" do you mean that it's not in the package's IUSE
at all, or that it is disabled?


On the sentence prior to the one you quoted I advised the OP to enable 
opengl globally. Therefore, if a package has it then it is already 
enabled. So obviously I meant if it doesn't have it at all.



by "enable *it*" do you mean enable opengl, or enable those other flags?


For the same reason stated above I can only be refering to those other 
flags. Those are the flags that control GL acceleration. If you want 
your system to use your GPU as much as possible then you need to enable 
at least one of them on all the packages that support it. If a package 
has more than one it's between you and portage which one you choose 
because some packages depend on one or the other. So I find it easier to 
enable opengl first and then work through the conflicts to enable the 
others as much as possible.


--

Fernando Rodriguez

Re: [gentoo-user] New AMD hardware. Still can't boot (but making progress).

[Daniel Frey]

On 04/28/2017 12:51 PM, Alan Mackenzie wrote:
> Hello, Mick and the whole World.
> 
>   # Load a linux kernel, passing the root filesystem and init process as 
> parameters
>   echo "Loading kernel..."
>   echo "grub_platform = " $grub_platform
>   gfxpayload=text # For debugging.
>   linux /vmlinuz-4.9.16-gentoo root=/dev/md127 #rootfstype=btrfs 
> init=/usr/lib/systemd/systemd
>   echo "just loaded kernel"
>   boot
> 
>   #echo "Loading initramfs..."
>   #initrd /initramfs-4.7.2.img
> }

Just FYI, I boot off of a RAID10 on imsm, albiet on older hardware. Some
time ago there was a change that the kernel would not autoassemble a
RAID with a kernel only, you need to use an initramfs. (OK, I just
looked and all I find are "my raid won't boot" results rather than the
changes I found a couple years ago. But I did try.)

I spent a couple of days trying to get mine to boot and eventually went
hell with it and used dracut to create an initramfs, and it was smooth
sailing after that.

Dan

Re: [gentoo-user] System borked after emerge -e world

[Mick]

On Thursday 27 Apr 2017 18:31:36 Dale wrote:

> I have it on a small post-it-note stuck on the side of my monitor.  I'm
> not sure, may have been Neil or Alan, that posted it this way.
> 
> Reboot Even If System Utterly Broken.

Another mnemonic:

B U S I E R 
<=<=<=<=<=<

Of course, you're meant to key it in backwards.

I think Alan mentioned it once or twice on this list.
-- 
Regards,
Mick

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

Re: [gentoo-user] How to get memtest onto a USB drive

[Dale]

Peter Humphrey wrote:
> Hello list,
>
> I have a nearly new machine which is already showing signs of hardware 
> failure. I'd like to check its memory, for which memtest86+ seems suitable. 
> But I can't install it via portage because this is a UEFI machine and so its 
> boot partition is vfat; the install phase fails because symbolic links can't 
> be done.
>
> I downloaded an ISO image from http://www.memtest.org/#downiso and used 
> unetbootin to write it to the USB. It won't boot. I then tried dd but got 
> the same result.
>
> What's the recommended way to put a bootable memtest86+ image, or 
> equivalent, on a USB stick?
>


This is different but I found this that says it works with your setup. 
I don't have one like yours, just googled for a possible solution. 

http://www.memtest86.com/technical.htm#usage 

If you think that will work, may want to try that memtest.  If not,
please ignore me.  ;-)

Dale

:-)  :-) 

Re: [gentoo-user] How to get memtest onto a USB drive

[Peter Humphrey]

On Friday 28 Apr 2017 22:33:35 Peter Humphrey wrote:
> Hello list,
> 
> I have a nearly new machine which is already showing signs of hardware
> failure. I'd like to check its memory, for which memtest86+ seems
> suitable. But I can't install it via portage because this is a UEFI
> machine and so its boot partition is vfat; the install phase fails
> because symbolic links can't be done.
> 
> I downloaded an ISO image from http://www.memtest.org/#downiso and used
> unetbootin to write it to the USB. It won't boot. I then tried dd but got
> the same result.
> 
> What's the recommended way to put a bootable memtest86+ image, or
> equivalent, on a USB stick?

Never mind. I dd'd the .bin file instead of the .iso and it did try to boot. 
Got nowhere though because the damn thing's jiggered.

Sorry about the noise.

-- 
Regards
Peter

[gentoo-user] How to get memtest onto a USB drive

[Peter Humphrey]

Hello list,

I have a nearly new machine which is already showing signs of hardware 
failure. I'd like to check its memory, for which memtest86+ seems suitable. 
But I can't install it via portage because this is a UEFI machine and so its 
boot partition is vfat; the install phase fails because symbolic links can't 
be done.

I downloaded an ISO image from http://www.memtest.org/#downiso and used 
unetbootin to write it to the USB. It won't boot. I then tried dd but got 
the same result.

What's the recommended way to put a bootable memtest86+ image, or 
equivalent, on a USB stick?

-- 
Regards
Peter

Re: [gentoo-user] New AMD hardware. Still can't boot (but making progress).

[J. Roeleveld]

On April 28, 2017 9:51:07 PM GMT+02:00, Alan Mackenzie  wrote:
>Hello, Mick and the whole World.
>
>On Tue, Apr 25, 2017 at 00:00:07 +0100, Mick wrote:
>
>> Have a look here in case it helps:
>
>> https://wiki.archlinux.org/index.php/GRUB#RAID
>
>In the end, I went with grub2, and it has taken a lot of effort to get
>not very far.  Grub's documentation is suboptimal.
>
>The state I've managed to get to is that grub appears to have loaded my
>kernel (it no longer gives an error message about it not being loaded),
>but then hangs without the kernel outputting even a single message.  At
>this point, only the reset-button will do anything - Ctrl-Alt-Del does
>nothing.
>
>Just for reference's sake, my boot setup is EFI on GPT.  I have two
>NVMe
>SSDs, which will be working together in several mdadm RAID pairs.  grub
>can read my SSDs, reporting correctly their partitioning and being able
>to cat the grub configuration file.
>
>So, why, after apparently loading the kernel, does grub fail to start
>it?
>Any ideas, anybody?
>
>Here are the relevant bits of my grub.cfg:
>
>#
>
># Example configuration for GRUB
># Much of this example configuration was taken from the GRUB manual.
>
># Menu timeout
>timeout=5
>
># Default menu entry
>default=0
>
># If we have a font available, start graphical output.
>if loadfont unifont; then
>   echo "Loading unifont"
>   # Output resolution for GRUB (eg. 1024x768 or 'auto').
>   gfxmode=auto
>
>   # Output resolution for Linux (VESAFB only).
>   # 'keep' means use the same resolution as GRUB.
>   # For other framebuffer drivers, pass a resolution using the video=
>kernel param.
>   gfxpayload=keep
>
>   # Load all video drivers.
>   insmod all_video
>
>   # Switch to graphical output.
>   terminal_output gfxterm
>fi
>
># Load modules necessary to find any boot files (/boot).
>
># Partition table(s).
>#insmod part_msdos
>insmod part_gpt
>
># Necessary for the root partition
>insmod mdraid1x
>insmod mdraid09
>
># Necessary for kernel messages to show
>insmod efi_gop
>insmod efi_uga
>insmod font
>if loadfont ${prefix}/fonts/unicode.pf2
>then
>   insmod gfxterm
>   set gfxmode=auto
>   set gfxpayload=keep
>   terminal_output gfxterm
>fi
>
>menuentry "Gentoo Linux 4.9.16-1" {
>   # Filesystem for /boot
>   #insmod btrfs
>   insmod fat
>   insmod exfat
>   insmod ext2
>   #insmod xfs
>   #insmod zfs
>
>   # Search all block devices for a matching UUID (for /boot)
>   #search --set=root --fs-uuid 33d4013a-ec25-4462-a540-8078aeb8ed17
>   #search --set=root (md/127)
>   #root=(md/127)
>   set root=(hd1,gpt2) # i.e. /boot
>   echo "Just set root"
>
>   # Load a linux kernel, passing the root filesystem and init process as
>parameters
>   echo "Loading kernel..."
>   echo "grub_platform = " $grub_platform
>   gfxpayload=text # For debugging.
>   linux /vmlinuz-4.9.16-gentoo root=/dev/md127 #rootfstype=btrfs
>init=/usr/lib/systemd/systemd
>   echo "just loaded kernel"
>   boot
>
>   #echo "Loading initramfs..."
>   #initrd /initramfs-4.7.2.img
>}
>
>#
>
>> -- 
>> Regards,
>> Mick

Have you tried connecting using ssh after boot?
Also, do you have the EFI console support in your kernel?

Not near my desktop to check which ones exactly.

--
Joost
-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Re: [gentoo-user] New AMD hardware. Still can't boot (but making progress).

[Alan Mackenzie]

Hello, Mick and the whole World.

On Tue, Apr 25, 2017 at 00:00:07 +0100, Mick wrote:

> Have a look here in case it helps:

> https://wiki.archlinux.org/index.php/GRUB#RAID

In the end, I went with grub2, and it has taken a lot of effort to get
not very far.  Grub's documentation is suboptimal.

The state I've managed to get to is that grub appears to have loaded my
kernel (it no longer gives an error message about it not being loaded),
but then hangs without the kernel outputting even a single message.  At
this point, only the reset-button will do anything - Ctrl-Alt-Del does
nothing.

Just for reference's sake, my boot setup is EFI on GPT.  I have two NVMe
SSDs, which will be working together in several mdadm RAID pairs.  grub
can read my SSDs, reporting correctly their partitioning and being able
to cat the grub configuration file.

So, why, after apparently loading the kernel, does grub fail to start it?
Any ideas, anybody?

Here are the relevant bits of my grub.cfg:

#

# Example configuration for GRUB
# Much of this example configuration was taken from the GRUB manual.

# Menu timeout
timeout=5

# Default menu entry
default=0

# If we have a font available, start graphical output.
if loadfont unifont; then
echo "Loading unifont"
# Output resolution for GRUB (eg. 1024x768 or 'auto').
gfxmode=auto

# Output resolution for Linux (VESAFB only).
# 'keep' means use the same resolution as GRUB.
# For other framebuffer drivers, pass a resolution using the video= 
kernel param.
gfxpayload=keep

# Load all video drivers.
insmod all_video

# Switch to graphical output.
terminal_output gfxterm
fi

# Load modules necessary to find any boot files (/boot).

# Partition table(s).
#insmod part_msdos
insmod part_gpt

# Necessary for the root partition
insmod mdraid1x
insmod mdraid09

# Necessary for kernel messages to show
insmod efi_gop
insmod efi_uga
insmod font
if loadfont ${prefix}/fonts/unicode.pf2
then
insmod gfxterm
set gfxmode=auto
set gfxpayload=keep
terminal_output gfxterm
fi

menuentry "Gentoo Linux 4.9.16-1" {
# Filesystem for /boot
#insmod btrfs
insmod fat
insmod exfat
insmod ext2
#insmod xfs
#insmod zfs

# Search all block devices for a matching UUID (for /boot)
#search --set=root --fs-uuid 33d4013a-ec25-4462-a540-8078aeb8ed17
#search --set=root (md/127)
#root=(md/127)
set root=(hd1,gpt2) # i.e. /boot
echo "Just set root"

# Load a linux kernel, passing the root filesystem and init process as 
parameters
echo "Loading kernel..."
echo "grub_platform = " $grub_platform
gfxpayload=text # For debugging.
linux /vmlinuz-4.9.16-gentoo root=/dev/md127 #rootfstype=btrfs 
init=/usr/lib/systemd/systemd
echo "just loaded kernel"
boot

#echo "Loading initramfs..."
#initrd /initramfs-4.7.2.img
}

#

> -- 
> Regards,
> Mick

-- 
Alan Mackenzie (Nuremberg, Germany).

[gentoo-user] Re: In search of an truecolor-capable terminal emulator

[Ian Zimmerman]

On 2017-04-28 10:10, Fernando Rodriguez wrote:

> No. I meant you can't enable them *all* globally, meaning opengl,
> gles, egl, etc. It's kind of the same situation as with GUI toolkits,
> you can't enable them all globally because some packages support more
> than one that you have to choose at compile time.

I think we need to spell out what you mean by this:

>>> if any of them have egl or gles and not opengl then enable it.

by "not having opengl" do you mean that it's not in the package's IUSE
at all, or that it is disabled?

by "enable *it*" do you mean enable opengl, or enable those other flags?

Sorry, I realize this may sound a bit pedantic and there may also be
some language barrier.  I often re-read my posts before sending them off
and edit them for clarity, even if this editing makes them look like
they have been written by a lawyer (which I am not).

-- 
Please *no* private Cc: on mailing lists and newsgroups
Personal signed mail: please _encrypt_ and sign
Don't clear-text sign:
http://primate.net/~itz/blog/the-problem-with-gpg-signatures.html

Re: [gentoo-user] Re: In search of an truecolor-capable terminal emulator

[Fernando Rodriguez]

On 04/26/2017 03:11 PM, Ian Zimmerman wrote:

On 2017-04-26 12:26, Fernando Rodriguez wrote:


My USE_FLAGS (make.conf) are:



USE="nvidia X lua sdl mp3 flac jack alsa gtk cairo sndfile
qt3support kpathsea gif tga jpeg png jpeg2k mad dvb dvdr encode lzo
bzip2 ogg sox v4l v4l2 vorbis x264 x265 zsh-completion -hal -lirc"



I would start by enabling opengl globally. Then look at the dependency
graph for the packages that you want to accelerate (especially the
toolkits and the whole x11 stack) and if any of them have egl or gles
and not opengl then enable it. You can't enable them all on make.conf
because sometimes they conflict.


Is this advice not in contradiction with itself?  Or what do you mean by
"enabling opengl globally" other than setting a USE flag in make.conf?



No. I meant you can't enable them *all* globally, meaning opengl, gles, 
egl, etc. It's kind of the same situation as with GUI toolkits, you 
can't enable them all globally because some packages support more than 
one that you have to choose at compile time.


--

Fernando Rodriguez

Re: [gentoo-user] gcc with graphite flag?

[Corbin Bird]

On 04/28/2017 02:08 AM, Walter Dnes wrote:
>   In the past, there used to be dire warnings about the difficulty of
> installing gcc with graphite.  Have things become easier with 5.4.0?
> Here's a "dry run" output...
> 
> 
> [i660][waltdnes][~] USE="graphite" emerge -pv gcc
> 
> These are the packages that would be merged, in order:
> 
> Calculating dependencies... done!
> [ebuild  N ] dev-libs/isl-0.15:0/15::gentoo  USE="-static-libs" 1,375 KiB
> [ebuild   R] sys-devel/gcc-5.4.0-r3:5.4.0::gentoo  USE="cxx fortran 
> graphite* nptl sanitize vtv (-altivec) (-awt) -cilk -debug -doc 
> (-fixed-point) -gcj -go (-hardened) (-jit) (-libssp) -mpx (-multilib) -nls 
> -nopie -nossp -objc -objc++ -objc-gc -openmp -regression-test -vanilla" 0 KiB
> 
> Total: 2 packages (1 new, 1 reinstall), Size of downloads: 1,375 KiB
> 
> 


On my Gentoo box ...

[ebuild   R] sys-devel/gcc-5.4.0-r3:5.4.0::gentoo  USE="cxx fortran
gcj graphite (multilib) nls nptl objc objc++ objc-gc openmp sanitize vtv
(-altivec) (-awt) -cilk -debug -doc (-fixed-point) -go (-hardened)
(-jit) (-libssp) -mpx -nopie -nossp -regression-test -vanilla" 0 KiB


"cloog" has been removed, going from GCC 4.9.4 -> GCC 5.4.0

No problems with "graphite" in either version of GCC, on my Gentoo box.

Corbin

[gentoo-user] Re: gcc with graphite flag?

[Nikos Chantziaras]

On 04/28/2017 10:08 AM, Walter Dnes wrote:

  In the past, there used to be dire warnings about the difficulty of
installing gcc with graphite.  Have things become easier with 5.4.0?


I don't remember any issues with it, even with 4.9.

Re: [gentoo-user] How do you manage manually compiled software?

[Daniel Campbell]

On 04/27/2017 07:19 PM, Michael Orlitzky wrote:
> On 04/27/2017 09:33 PM, Danny YUE wrote:
>> Hi guys,
>>
>> I am compiling RISC-V tools...I am just curious how do you manage your
>> manually compiled software?
> 
> Don't, write an ebuild for it.
> 
> 
> 
+1 for this. Even if it's a somewhat amateur ebuild, at least Portage
can manage it.

-- 
Daniel Campbell - Gentoo Developer
OpenPGP Key: 0x1EA055D6 @ hkp://keys.gnupg.net
fpr: AE03 9064 AE00 053C 270C  1DE4 6F7A 9091 1EA0 55D6



signature.asc
Description: OpenPGP digital signature

Re: [gentoo-user] How do you manage manually compiled software?

[Neil Bothwick]

On Thu, 27 Apr 2017 22:19:03 -0400, Michael Orlitzky wrote:

> > I am compiling RISC-V tools...I am just curious how do you manage your
> > manually compiled software?  
> 
> Don't, write an ebuild for it.

After checking bgo to see if anyone else has already invented that
particular wheel.


-- 
Neil Bothwick

Ohnosecond: That minuscule fraction of time in which you realize
you've just made a big mistake


pgpgFhU_mwhl0.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] How do you manage manually compiled software?

[Danny YUE]

On 2017-04-28 02:19, Michael Orlitzky  wrote:
> On 04/27/2017 09:33 PM, Danny YUE wrote:
>> Hi guys,
>> 
>> I am compiling RISC-V tools...I am just curious how do you manage your
>> manually compiled software?
>
> Don't, write an ebuild for it.

Well, since everyone is recommending the "correct Gentoo way"..
I will give it a try.

Thanks.

[gentoo-user] gcc with graphite flag?

[Walter Dnes]

  In the past, there used to be dire warnings about the difficulty of
installing gcc with graphite.  Have things become easier with 5.4.0?
Here's a "dry run" output...


[i660][waltdnes][~] USE="graphite" emerge -pv gcc

These are the packages that would be merged, in order:

Calculating dependencies... done!
[ebuild  N ] dev-libs/isl-0.15:0/15::gentoo  USE="-static-libs" 1,375 KiB
[ebuild   R] sys-devel/gcc-5.4.0-r3:5.4.0::gentoo  USE="cxx fortran 
graphite* nptl sanitize vtv (-altivec) (-awt) -cilk -debug -doc (-fixed-point) 
-gcj -go (-hardened) (-jit) (-libssp) -mpx (-multilib) -nls -nopie -nossp -objc 
-objc++ -objc-gc -openmp -regression-test -vanilla" 0 KiB

Total: 2 packages (1 new, 1 reinstall), Size of downloads: 1,375 KiB


-- 
Walter Dnes 
I don't run "desktop environments"; I run useful applications