Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP

Loading…

sys-kernel/geek-sources notes [not a bug] #21

Open
init6 opened this Issue · 36 comments

6 participants

@init6
Owner

sys-kernel/geek-sources-3.10.7
USE gentoo, fedora, suse contain a lot of bugs!
USE grsec with 4420_grsecurity-2.9.1-3.10.7-201308182203.patch same build error same bug and almost same fix

# cat Makefile.patch
diff -urN a/Makefile b/Makefile
--- a/Makefile  2013-08-20 09:13:23.000000000 +0300
+++ b/Makefile  2013-08-20 12:17:31.424283049 +0300
@@ -793,7 +793,7 @@


ifeq ($(KBUILD_EXTMOD),)
-core-y     += kernel/ mm/ spl/ fs/ ipc/ security/ crypto/ block/
+core-y     += kernel/ mm/ spl/ fs/ ipc/ security/ crypto/ block/ grsecurity/

vmlinux-dirs    := $(patsubst %/,%,$(filter %/, $(init-y) $(init-m) \
            $(core-y) $(core-m) $(drivers-y) $(drivers-m) \
@init6
Owner

sys-kernel/geek-sources-3.10.9
need same fix for grsec
i`m not test USE gentoo, fedora, suse

@init6
Owner

For =sys-kernel/geek-sources-3.11.1 USE="grsec" need same Makefile.patch ^

@init6
Owner

=sys-kernel/geek-sources-3.11.5 with USE="grsec" for my not build fine.

@init6
Owner

For 3.11.5 with USE="grsec" need patch

cat grsec-3.11.5.patch 
diff -urN a/linux-3.11.5-geek/ipc/shm.c b/linux-3.11.5-geek/ipc/shm.c
--- a/linux-3.11.5-geek/ipc/shm.c   2013-10-18 17:13:11.610986558 +0300
+++ b/linux-3.11.5-geek/ipc/shm.c   2013-10-18 17:05:29.000000000 +0300
@@ -1105,7 +1105,7 @@
     * additional creator id...
     */
    ns = current->nsproxy->ipc_ns;
-   shp = shm_lock_check(ns, shmid);
+   /* shp = shm_lock_check(ns, shmid); */
    if (IS_ERR(shp)) {
        err = PTR_ERR(shp);
        goto out;
@init6
Owner

with 3.11.5 hardened everything is very bad :( with patch kernel build fine. but does not boot.

@init6
Owner

Why no any updates? Now it is a reality.

My five-year-old laptop Acer Aspire 5730ZG is starting to die. Until now, it was just minor damage. Now his repair a serious blow to my wallet. I simply do not have the money to buy new hardware. And besides, I do not have a permanent job and a regular source of income. Until I do not solve these problems I stop the activity of all my projects.

Over the overlay agreed to watch Pinkbyte, deterok, Yamakuzure, grondinm, CarelessChaser, tazhate.

Почему нет обновлений? Теперь это уже действительность.

Мой пятилетней давности ноутбук Acer Aspire 5730ZG начинает умирать. До сих пор это были просто незначительные поломки. Теперь его починка серьезный удар по моему бумажнику. У меня попросту нет денег на покупку нового железа. И к тому же у меня нет постоянной работы и постоянного источника доходов. До тех пор пока я не разрешу эти проблемы я прекращаю свою деятельность над всеми моими проектами.

За оверлеем согласились присматривать Pinkbyte, deterok, Yamakuzure, grondinm, CarelessChaser, tazhate.

Всю новую информацию я буду размещать тут. Оставайтесь с нами. „И спасибо за рыбу.“

@grondinm
Collaborator

This is sad news indeed. Hope everything works out for you and thank you for all your hard work.

@init6
Owner

If you want to work directly on the overlay, leave your requests below. And I add you to coworker in this overlay!

@Yamakuzure
Collaborator

I'd like to be added. As you can see, I have the same username here as in the gentoo forums. ;)

@init6
Owner

I'd like to be added. As you can see, I have the same username here as in the gentoo forums. ;)

Yamakuzure Done!

This was referenced
@grondinm
Collaborator
@init6
Owner

so if the offer still stands i would like to be added.

grondinm Done!

@grondinm
Collaborator

Thank you

@init6
Owner

I added some new information to the wiki about geek-sources and donate.

At me while without much change.

I want to try start solving issue #29. All my offline work I will upload exactly a month - 1 th of January! And Thank you all for your work.

@mgomersbach

I also use geek-sources for some projects which will be maintained. Mainly PF patches and the odd ZFS and AUFS (Maitreya on the forums). As I will fork it and try to deliver back upstream here.

@init6
Owner

I also use geek-sources for some projects which will be maintained. Mainly PF patches and the odd ZFS and AUFS (Maitreya on the forums). As I will fork it and try to deliver back upstream here.

You're Welcome!

@init6
Owner

Less than two weeks these {USE flags/eclasses} will be removed: geek-ubuntu.eclass, geek-xenomai.eclass

These {USE flags/eclasses} change their names: old geek-grsec.eclass -> new geek-hardened.eclass

New geek-grsec.eclass will only contain vanilla GrSecurity patch.

About others updates will learn later.

And pay attention to this moment github and "This comparison is taking too long to generate" i fix only geek-ice.eclass If you need the rest of these eclasses start to think about how to solve this problem. In case of errors, I just turn off the corresponding USE flag.

@init6
Owner

As I promised earlier

All my offline work I will upload exactly a month - 1 th of January! And Thank you all for your work.

Now with this commit:

  • Removed {USE flags/eclasses} geek-ubuntu.eclass, geek-xenomai.eclass

  • Add {USE flags/eclasses} geek-hardened.eclass, geek-openvz.eclass, geek-openwrt.eclass and geek-vars.eclass for internal-used variables.

  • Old USE=grsec now apply only vanilla GrSecurity patch. For full hardened-sources patchset there is new USE=hardened.

  • Almost in all eclasses there is new feature called ApplyUserPatch Example In this case it is needed for apply a any patch from the "/etc/portage/patches/sys-kernel/geek-sources/aufs" immediately after applying the patches aufs itself. Almost like epatch_user but for the current patchset. In this case for aufs.

  • Issue #28 now fxd.

  • And many other changes and optimizations… For example now almost all cp/mv are replaced by copy or move from geek-utils.eclass To see all changes use git diff

I draw your attention that Issue #39 fxd only for geek-ice.eclass

Merry Christmas and Happy New Year!

@init6
Owner

With one of the following commits you get the opportunity to build sys-kernel/geek-sources based not only on vanilla linux kernel sources but also based on Red-Hat Enterprise Linux kernel sources (The same USE="rh". Supported version rhel6 - kernel-2.6.32-19.el6.src.rpm or beta rhel7 - kernel-3.10.0-54.0.1.el7.src.rpm) or Oracle’s Unbreakable Enterprise Linux kernel sources (New USE="uek". Supported version now only kernel-uek-2.6.32-400.34.3.el6uek.src.rpm because 3.8 not in overlay)

Stay tuned.

@grondinm
Collaborator

so i'm running into issues merging geek-sources...i'm not sure what it causing it but it keeps skipping the update to latest upstream. so for example if i try to merge 3.10.32 i end up with 3.10.0...i have looked at the latest commits and nothing seems to be responsible that is why i am confused...

@init6
Owner

And no any useful information…

grondinm all patches/patchsets except:

  • lqx
  • rh
  • pf
  • zen
  • and from now + uek

depends on linux kernel sources (for example 3.13.0) + latest upstream updates (for example 3.13.5). But those USE flags that listed above provide their own upstream updates. And patches/patchsets listed above it is possible to apply only on "clean" kernel without latest upstream updates (for example 3.13.0)

If you want to change something that I personally have nothing against.

But where and what to change I hope that clear?

And by the way all this crap shown in this understandable picture in wiki…

@grondinm
Collaborator

sorry i should have provided more info. i do not use any of those patches tho. only ck bfq brand and optimize...and it just worked fine on 3.12 to 3.12.13 with same patches....adds to confusion...

@init6
Owner

grondinm then you change GEEKSOURCES_PATCHING_ORDER and missed something?

Or is it a new bug. However, again, just comprehensive information even does not allow answer the question about what it might be caused.

And btw for my all fine.

@init6
Owner

I need some statistics so Which patches do you use in geek-sources?

@init6
Owner

From 26.04.2014 my work with overlay "init6" and "sys-kernel/geek-sources" officially stopped. Luks disk partition with gentoo completely destroyed and will not be restored. And besides no much sense of restoring the system on actually dead hardware.

If you want to say something, say it.
If you want to do something, do it.
If you want something to help - do it.

Almost the same text but in Russian and google+... Почти тот же текст но на русском языке и на google+

@init6
Owner

+ tazhate Welcome!

@init6
Owner

2 @nilekurt :

Now I need to fix all this ^ so @nilekurt Please do not do more of this. Thx.

@nilekurt
Collaborator

@init6:

  • Noted, but seeing as BFS is included in the ck-patchset and CK just hasn't gotten around to compiling ck2 yet, I thought that it was only appropriate. Edit: To clarify, the BFS variable is only used in cases such as this, when CK has a patchset out, but hasn't yet updated it to contain the newest version of BFS.
  • Understood. I'll see what I can do about that, by the way.
  • I was unaware about any of my changes directly affecting this. Can you point out where I went wrong?
  • I've been careful not to touch any unrelated USE-flags, but accidents do happen. Can you point to a specific commit?
@init6
Owner

@nilekurt I already fix all. And I need only to check. But it needs download all git/svn repo and it will require a lot of time…

My untested fixes will soon be in the overlay.

@init6
Owner

@nilekurt I fix it. Please dont do that again.

@bigfreshy

https://aur.archlinux.org/packages/linux-bld/?setlang=sk
new bld up to 3.14.11, i'll see if i can ge this working

@grondinm
Collaborator

@init6 i had left funtoo/gentoo for a while...i am now back to it and just wanted to check with you as i still have write access would you rather i make any updates(pretty much only kernel version bumps) myself or fork and submit pull requests?

@init6
Owner

@grondinm everything is still the same as it is written there. And you have the right to work directly with overlay.

@init6
Owner

I make Yet another gentoo fork What do you think about this?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.