Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
On Sat, 10 Dec 2022 at 20:11, Sebastian Kuzminsky wrote: > As Dewey reported later, procfs_macros.h is used in the RTAI backend of > RTAPI, and deleting it breaks the RTAI build unless we turn off our > /proc code at the same time. Not for me, unless I have messed up somehow. (Not unlikely) I deleted the file in the 2.9 branch, make clean, ./autogen.sh, ./configure, make, sudo make setuid, and the build went fine. runtests then failed with a stack trace, but I don't know if that is related yet. This is RTAI running on a VM, on a Mac. There is some weirdness. --- Dec 21 20:10:38 buster kernel: [308421.298377] I-pipe domain: Linux Dec 21 20:10:38 buster kernel: [308421.298377] RIP: 0010:__ipipe_halt_root+0x2e/0x40 Dec 21 20:10:38 buster kernel: [308421.298378] Code: 40 c1 01 00 65 48 8b 15 c8 71 fe 7e 48 01 d0 48 0f ba 30 00 48 83 78 08 00 75 10 85 ff 74 09 31 c0 89 c1 fb 0f 01 c9 c3 fb f4 e8 4c 37 09 00 fb c3 66 2e 0f 1f 84 00 00 00 00 00 48 89 fa 8b Dec 21 20:10:38 buster kernel: [308421.298378] RSP: 0018:81c03ed8 EFLAGS: 0246 Dec 21 20:10:38 buster kernel: [308421.298379] RAX: 888135a1c140 RBX: RCX: 0246 Dec 21 20:10:38 buster kernel: [308421.298379] RDX: 888135a0 RSI: RDI: Dec 21 20:10:38 buster kernel: [308421.298380] RBP: 81c59d68 R08: 888135a0 R09: 000303c0 Dec 21 20:10:38 buster kernel: [308421.298389] R10: 0003498c1108d477 R11: 0001 R12: 81c0e480 Dec 21 20:10:38 buster kernel: [308421.298389] R13: 888139ff8140 R14: R15: Dec 21 20:10:38 buster kernel: [308421.298390] FS: () GS:888135a0() knlGS: Dec 21 20:10:38 buster kernel: [308421.298390] CS: 0010 DS: ES: CR0: 80050033 Dec 21 20:10:38 buster kernel: [308421.298390] CR2: 7fcf433a7006 CR3: 01c08005 CR4: 001606f0 Dec 21 20:10:38 buster kernel: [308421.298391] Call Trace: Dec 21 20:10:38 buster kernel: [308421.298391] ? default_idle+0xe/0x20 Dec 21 20:10:38 buster kernel: [308421.298391] ? do_idle+0xe0/0x110 Dec 21 20:10:38 buster kernel: [308421.298392] ? cpu_startup_entry+0x5a/0x60 Dec 21 20:10:38 buster kernel: [308421.298392] ? start_kernel+0x450/0x45e Dec 21 20:10:38 buster kernel: [308421.298393] ? secondary_startup_64+0xa4/0xb0 Dec 21 20:10:38 buster kernel: [308421.299385] rcu: rcu_sched kthread starved for 15912 jiffies! g29725 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=0 Dec 21 20:10:38 buster kernel: [308421.299386] rcu: RCU grace-period kthread stack dump: Dec 21 20:10:38 buster kernel: [308421.299387] rcu_sched I0 11 2 0x8000 Dec 21 20:10:38 buster kernel: [308421.299388] Call Trace: Dec 21 20:10:38 buster kernel: [308421.299392] ? context_switch+0x52/0x250 Dec 21 20:10:38 buster kernel: [308421.299395] ? __schedule+0x12e/0x430 Dec 21 20:10:38 buster kernel: [308421.299396] ? schedule+0x61/0x90 Dec 21 20:10:38 buster kernel: [308421.299397] ? schedule_timeout+0x147/0x2c0 Dec 21 20:10:38 buster kernel: [308421.299398] ? __prepare_to_swait+0x46/0x70 Dec 21 20:10:38 buster kernel: [308421.299400] ? __next_timer_interrupt+0xc0/0xc0 Dec 21 20:10:38 buster kernel: [308421.299402] ? rcu_gp_kthread+0x451/0x900 Dec 21 20:10:38 buster kernel: [308421.299403] ? wait_rcu_exp_gp+0x30/0x30 Dec 21 20:10:38 buster kernel: [308421.299405] ? kthread+0x11e/0x140 Dec 21 20:10:38 buster kernel: [308421.299406] ? kthread_create_worker_on_cpu+0x40/0x40 Dec 21 20:10:38 buster kernel: [308421.299407] ? ret_from_fork+0x20/0x30 -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
On Wed, 21 Dec 2022 at 19:29, andy pugh wrote: > In fact I see _nothing_ in /proc for rtapi (with 2.8 RTAI, 2.9 RTAI > and 2.8.4 preempt-rt (from the ISO + update) Possibly worth noting that this is with kernel 4.19.195-rtai-amd64 (from our repository) -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
On Sat, 10 Dec 2022 at 20:11, Sebastian Kuzminsky wrote: > There are a bunch of things in /proc/rtapi that seem useful, that use > procfs_macros.h. Unfortunately reading from any of them produces > garbage. That seems not to be the case for me: In fact I see _nothing_ in /proc for rtapi (with 2.8 RTAI, 2.9 RTAI and 2.8.4 preempt-rt (from the ISO + update) With the RTAI versions I do see a /proc/rtai but that isn't ours (I don't think) and doesn't contain the "status" that you mention trying. -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
On 12/10/22 08:14, Andy Pugh wrote: On 10 Dec 2022, at 14:49, Dewey Garrett wrote: nada -- this file is intree but not used Interesting. Perhaps we can just delete it. As Dewey reported later, procfs_macros.h is used in the RTAI backend of RTAPI, and deleting it breaks the RTAI build unless we turn off our /proc code at the same time. I'm testing 2.8 on Wheezy with the 3.4-9-rtai-686-pae kernel from our package archive. There are a bunch of things in /proc/rtapi that seem useful, that use procfs_macros.h. Unfortunately reading from any of them produces garbage. Writing to /proc/rtapi/debug works fine and does the expected thing (rtapi messages start showing up in the kernel log). This is representative: ~~~ $ cat /proc/rtapi/status ssage level = 5 975 nSec ~~~ It seems like `seq_file` may be a better way for us to supply the contents of our files in /proc? It's part of the mainline kernel. -- Sebastian Kuzminsky ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
> On 10 Dec 2022, at 15:08, Dewey Garrett wrote: > > Macros used to implement the /proc interface Probably redundant since we stopped being able to set debug level in rtapi with echo 7 > /proc/debug ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
Great! Thanks for figuring this out Dewey :-) On December 10, 2022 10:43:06 AM MST, Dewey Garrett wrote: >In <863c8874-ef3b-4f5d-a1fb-6135dc438...@highlab.com> Sebastian Kuzminsky > writes: > >>Does it build on both uspace and RTAI? > >$ git status -uno >On branch master >Your branch is up to date with 'origin/master'. > >Changes to be committed: > (use "git restore --staged ..." to unstage) > modified: src/rtapi/README > deleted:src/rtapi/procfs_macros.h > deleted:src/rtapi/rtapi_proc.h > >$ git diff -b --cached src/rtapi/README >diff --git a/src/rtapi/README b/src/rtapi/README >index 3ae1f121ac..43a894 100644 >--- a/src/rtapi/README >+++ b/src/rtapi/README >@@ -10,7 +10,6 @@ README : this file > rtapi.h : decls for the generic RTAPI functions. > rtapi_app.h : decls for the kernel modules > rtapi_common.h : Collection of typedefs, etc, used internally by RTAPI. >-procfs_macros.h : Macros used to implement the /proc interface > rtapi_proc.h: more stuff implementing the /proc interface > rtai_rtapi.c: implementation of real-time API, for RTAI > rtai_ulapi.c: implementation of user-level API, for RTAI > >#=== >Tested: $ make clean && make && sudo make setuid > >builds ok on both uspace and rtai (rtai in a virtualbox) > >Ref: >1) uspace >$ uname -a >Linux beye 5.10.78-rt55 #4 SMP PREEMPT_RT Fri Dec 2 16:47:27 MST 2022 x86_64 >GNU/Linux >$ lsb_release -a >No LSB modules are available. >Distributor ID:Debian >Description: Debian GNU/Linux 11 (bullseye) >Release: 11 >Codename: bullseye > >2) rtai (virtualbox): >$ uname -a >Linux dvm 4.19.195-rtai-amd64 #5 SMP PREEMPT Sun Jul 11 19:13:27 BST 2021 >x86_64 GNU/Linux >$ lsb_release -a >No LSB modules are available. >Distributor ID:Debian >Description: Debian GNU/Linux 10 (buster) >Release: 10 >Codename: buster > >#=== >runtests passed on both systems above > >Note: also tested uspace on non-std kernel for bullseye (6.0.5-rt14) >$ uname -a >Linux beye 6.0.5-rt14 #1 SMP PREEMPT_RT Sat Dec 3 07:07:29 MST 2022 x86_64 >GNU/Linux > >-- >Dewey Garrett > > > >___ >Emc-developers mailing list >Emc-developers@lists.sourceforge.net >https://lists.sourceforge.net/lists/listinfo/emc-developers -- Sebastian Kuzminsky ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
Correction: for rtai, src/rtapi/rtapi_proc.h was not deleted, just changed: $ git diff -b --cached src/rtapi/rtapi_proc.h diff --git a/src/rtapi/rtapi_proc.h b/src/rtapi/rtapi_proc.h index dad9a7d57b..72d0c495f3 100644 --- a/src/rtapi/rtapi_proc.h +++ b/src/rtapi/rtapi_proc.h @@ -76,10 +76,6 @@ */ #if defined( CONFIG_PROC_FS ) && LINUX_VERSION_CODE < KERNEL_VERSION(3,10,0) -#define RTAPI_USE_PROCFS - -#include "procfs_macros.h" /* macros for read functions */ - struct proc_dir_entry *rtapi_dir = 0; /* /proc/rtapi directory */ static struct proc_dir_entry *status_file = 0; /* /proc/rtapi/status */ static struct proc_dir_entry *modules_file = 0;/* /proc/rtapi/modules * / #= RTAPI_USE_PROCFS is used in one file (2 places): $ find . -iname '*.*[ch]' -exec grep -Hn RTAPI_USE_PROCFS {} \; ./rtapi/rtai_rtapi.c:189:#ifdef RTAPI_USE_PROCFS ./rtapi/rtai_rtapi.c:273:#ifdef RTAPI_USE_PROCFS -- Dewey Garrett ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
In <863c8874-ef3b-4f5d-a1fb-6135dc438...@highlab.com> Sebastian Kuzminsky writes: >Does it build on both uspace and RTAI? $ git status -uno On branch master Your branch is up to date with 'origin/master'. Changes to be committed: (use "git restore --staged ..." to unstage) modified: src/rtapi/README deleted:src/rtapi/procfs_macros.h deleted:src/rtapi/rtapi_proc.h $ git diff -b --cached src/rtapi/README diff --git a/src/rtapi/README b/src/rtapi/README index 3ae1f121ac..43a894 100644 --- a/src/rtapi/README +++ b/src/rtapi/README @@ -10,7 +10,6 @@ README : this file rtapi.h : decls for the generic RTAPI functions. rtapi_app.h : decls for the kernel modules rtapi_common.h : Collection of typedefs, etc, used internally by RTAPI. -procfs_macros.h : Macros used to implement the /proc interface rtapi_proc.h: more stuff implementing the /proc interface rtai_rtapi.c: implementation of real-time API, for RTAI rtai_ulapi.c: implementation of user-level API, for RTAI #=== Tested: $ make clean && make && sudo make setuid builds ok on both uspace and rtai (rtai in a virtualbox) Ref: 1) uspace $ uname -a Linux beye 5.10.78-rt55 #4 SMP PREEMPT_RT Fri Dec 2 16:47:27 MST 2022 x86_64 GNU/Linux $ lsb_release -a No LSB modules are available. Distributor ID: Debian Description:Debian GNU/Linux 11 (bullseye) Release:11 Codename: bullseye 2) rtai (virtualbox): $ uname -a Linux dvm 4.19.195-rtai-amd64 #5 SMP PREEMPT Sun Jul 11 19:13:27 BST 2021 x86_64 GNU/Linux $ lsb_release -a No LSB modules are available. Distributor ID: Debian Description:Debian GNU/Linux 10 (buster) Release:10 Codename: buster #=== runtests passed on both systems above Note: also tested uspace on non-std kernel for bullseye (6.0.5-rt14) $ uname -a Linux beye 6.0.5-rt14 #1 SMP PREEMPT_RT Sat Dec 3 07:07:29 MST 2022 x86_64 GNU/Linux -- Dewey Garrett ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
Does it build on both uspace and RTAI? On December 10, 2022 8:04:19 AM MST, Dewey Garrett wrote: >In Dewey Garrett writes: > > > >>nada -- this file is intree but not used >not used in any .c,.cc files but: > >$ find . -type f -exec grep -H procfs_macros {} \; >./src/rtapi/rtapi_proc.h:#include "procfs_macros.h"/* macros for read >functions */ >./src/rtapi/README:procfs_macros.h : Macros used to implement the /proc >interface > >However, when commenting out those lines, build still works >-- >Dewey Garrett > > > >___ >Emc-developers mailing list >Emc-developers@lists.sourceforge.net >https://lists.sourceforge.net/lists/listinfo/emc-developers -- Sebastian Kuzminsky ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
> On 10 Dec 2022, at 14:49, Dewey Garrett wrote: > > nada -- this file is intree but not used Interesting. Perhaps we can just delete it. But I fear this is just an opening salvo. ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
In Dewey Garrett writes: >nada -- this file is intree but not used not used in any .c,.cc files but: $ find . -type f -exec grep -H procfs_macros {} \; ./src/rtapi/rtapi_proc.h:#include "procfs_macros.h" /* macros for read functions */ ./src/rtapi/README:procfs_macros.h : Macros used to implement the /proc interface However, when commenting out those lines, build still works -- Dewey Garrett ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
$ pwd /home/git/linuxcnc-dev $ find . -iname procfs_macros.h ./src/rtapi/procfs_macros.h $ find . -type f -iname '*.*c' -exec grep -H procfs_macros {} \; nada -- this file is intree but not used -- Dewey Garrett ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] Fwd: Bug#1025433: Copyright issue
If some copyright is wrong or incomplete then we can just fix it and close that bug report? > Gesendet: Samstag, 10. Dezember 2022 um 00:40 Uhr > Von: "andy pugh" > An: "EMC developers" > Betreff: [Emc-developers] Fwd: Bug#1025433: Copyright issue > > For some reason this was not sent to emc-developers. > > Clearly, for reasons of his own this chap does not want LinuxCNC in Debian. > > Frankly, I am OK with this. We haven't been in before and all my machine > tools work just fine. > > This is not a fight I want to fight. > > -- Forwarded message - > From: Adam Ant > Date: Fri, 9 Dec 2022 at 23:23 > Subject: Re: [Emc-developers] Bug#1025433: Copyright issue > To: andy pugh > Cc: <1025...@bugs.debian.org> > > > > Source: linuxcnc > Severity serious > > > > Sent: Thursday, December 08, 2022 at 1:36 AM > From: "andy pugh" > To: "EMC developers" , "Adam Ant" < > adam...@engineer.com> > Subject: Re: [Emc-developers] Bug#1025433: Copyright issue > > > > On Thu, 8 Dec 2022 at 00:42, Adam Ant adam...@engineer.com]> wrote: > > > > The correct course of action is to ask Paolo Mantegazza rather than > debating symantics. > You can not just pull substantial chunks of code from one source and then > claim that you wrote it. > But Paulo didn't write it _either_ > > If you go back in time to 2003: > https://github.com/LinuxCNC/linuxcnc/blob/07bc1e161f834d8b192fe279819261294e5fe150/src/rtapi/procfs_macros.h[https://github.com/LinuxCNC/linuxcnc/blob/07bc1e161f834d8b192fe279819261294e5fe150/src/rtapi/procfs_macros.h] > and compare that to a 1999 version of the RTAI file, the only line in > common is "#include " > > At some point _both_ files included a chunk of work from Erwin Rol: > > // proc print macros - Contributed by: Erwin Rol (er...@muffin.org[mailto: > er...@muffin.org]) > > _Most_ of both files was written by Erwin Rol. But both files existed > previously, and separately, as far as I can tell looking back in both > archives. > > > Regardless of who wrote the original code, you can not take a file, hack it > a little and then claim sole copyright yourself. > It brings the Debian project in to disrepute if they host the package. > > > -- > atp > "A motorcycle is a bicycle with a pandemonium attachment and is designed > for the especial use of mechanical geniuses, daredevils and lunatics." > — George Fitch, Atlanta Constitution Newspaper, 1912 > > ___ > Emc-developers mailing list > Emc-developers@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-developers > ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers