[gentoo-musl] plan9port build fails on multiple levels.

[gentoo-musl] plan9port build fails on multiple levels.

[Rendov Norra]

It fails at building its thread library, it fails at building
lib9/dirread, and I'm not sure if this is related at all but egrep
segfaults and kills the build.

[gentoo-musl] Re: plan9port build fails on multiple levels.

[Rendov Norra]

I should also note that this is on ARM

On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
> It fails at building its thread library, it fails at building
> lib9/dirread, and I'm not sure if this is related at all but egrep
> segfaults and kills the build.
>

[gentoo-musl] Re: plan9port build fails on multiple levels.

[Rendov Norra]

I have patches to fix the first two issues (or at least they compile),
but they're very hacky as I'm not an experienced coder.

On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
> I should also note that this is on ARM
>
> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>> It fails at building its thread library, it fails at building
>> lib9/dirread, and I'm not sure if this is related at all but egrep
>> segfaults and kills the build.
>>
>

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[parazyd]

Do share them. plan9port doesn't compile on amd64 either.

On Thu, 29 Jun 2017, Rendov Norra wrote:

> I have patches to fix the first two issues (or at least they compile),
> but they're very hacky as I'm not an experienced coder.
> 
> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
> > I should also note that this is on ARM
> >
> > On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
> >> It fails at building its thread library, it fails at building
> >> lib9/dirread, and I'm not sure if this is related at all but egrep
> >> segfaults and kills the build.
> >>
> >
> 

-- 
~ parazyd
GPG: 0333 7671 FDE7 5BB6 A85E  C91F B876 CB44 FA1B 0274

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[Anthony G. Basile]

On 6/30/17 3:02 AM, parazyd@dyne.org wrote:
> Do share them. plan9port doesn't compile on amd64 either.
>
> On Thu, 29 Jun 2017, Rendov Norra wrote:
>
>> I have patches to fix the first two issues (or at least they compile),
>> but they're very hacky as I'm not an experienced coder.

Excellent, ignore my last message.  If you open a bug report at
bugs.gentoo.org, describe the problem and upload the patches and I'll
put it on the tree.  I also maintain plan9port so I can fix it directly.

Having said that, I'm curious why people are interested in plan9port?

>>
>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>> I should also note that this is on ARM
>>>
>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>> It fails at building its thread library, it fails at building
>>>> lib9/dirread, and I'm not sure if this is related at all but egrep
>>>> segfaults and kills the build.
>>>>

-- 
Anthony G. Basile, Ph.D.
Gentoo Linux Developer [Hardened]
E-Mail    : basile@freeharbor.net
GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
GnuPG ID  : F52D4BBA

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[Rendov Norra]

[-- Attachment #1: Type: text/plain, Size: 1745 bytes --]

I've attached the patches here, but I hesitate to present them as a
solution since I'm incapable of testing them because of the segfaults
I'm experiencing with egrep.

I'll also note that the version
(e78ed7a67bbaa37dc57dfb37219f505c48755e14) is 3 years old. One big fix
that happened since then is the change of some files' extensions for
.C to .c, since they don't compile as .C .

As for why I'm interested in p9p, it's a feasible replacement base
system in place of GNU--being small but still functional--, and sam
and acme are useful editors.

On 6/30/17, Anthony G. Basile <basile@freeharbor.net> wrote:
> On 6/30/17 3:02 AM, parazyd@dyne.org wrote:
>> Do share them. plan9port doesn't compile on amd64 either.
>>
>> On Thu, 29 Jun 2017, Rendov Norra wrote:
>>
>>> I have patches to fix the first two issues (or at least they compile),
>>> but they're very hacky as I'm not an experienced coder.
>
> Excellent, ignore my last message.  If you open a bug report at
> bugs.gentoo.org, describe the problem and upload the patches and I'll
> put it on the tree.  I also maintain plan9port so I can fix it directly.
>
> Having said that, I'm curious why people are interested in plan9port?
>
>>>
>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>> I should also note that this is on ARM
>>>>
>>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>>> It fails at building its thread library, it fails at building
>>>>> lib9/dirread, and I'm not sure if this is related at all but egrep
>>>>> segfaults and kills the build.
>>>>>
>
> --
> Anthony G. Basile, Ph.D.
> Gentoo Linux Developer [Hardened]
> E-Mail    : basile@freeharbor.net
> GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
> GnuPG ID  : F52D4BBA
>
>
>

[-- Attachment #2: 1.patch --]
[-- Type: text/x-patch, Size: 236 bytes --]

--- /a/src/libthread/threadimpl.h	2014-03-05 21:43:20.000000000 +0000
+++ /b/src/libthread/threadimpl.h	2017-06-29 03:37:52.410011050 +0000
@@ -1,3 +1,4 @@
+#define _GNU_SOURCE
 #include "u.h"
 #include <errno.h>
 #include <sys/time.h>

[-- Attachment #3: 2.patch --]
[-- Type: text/x-patch, Size: 390 bytes --]

--- /a/src/libthread/Linux.c	2014-03-05 21:43:20.000000000 +0000
+++ /b/src/libthread/Linux.c	2017-06-29 18:17:47.550001078 +0000
@@ -1,4 +1,5 @@
 #include "threadimpl.h"
+#include <ucontext.h>
 
 #undef exits
 #undef _exits
@@ -435,7 +436,7 @@
 {
 	_exit(0);
 }
-
+/*
 #ifdef __arm__
 void
 makecontext(ucontext_t *uc, void (*fn)(void), int argc, ...)
@@ -461,3 +462,4 @@
 }
 #endif
 
+*/

[-- Attachment #4: 3.patch --]
[-- Type: text/x-patch, Size: 374 bytes --]

--- src/lib9/dirread.c	2014-03-05 21:43:20.000000000 +0000
+++ src/lib9/dirread.c.2	2017-06-29 19:49:19.780003697 +0000
@@ -15,7 +15,7 @@
 
 	/* This doesn't match the man page, but it works in Debian with a 2.2 kernel */
 	off = p9seek(fd, 0, 1);
-	nn = getdirentries(fd, (void*)buf, n, &off);
+	nn = getdents(fd, (void*)buf, n);
 	return nn;
 }
 #elif defined(__APPLE__) 

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[Anthony G. Basile]

On 6/30/17 5:14 PM, Rendov Norra wrote:
> I've attached the patches here, but I hesitate to present them as a
> solution since I'm incapable of testing them because of the segfaults
> I'm experiencing with egrep.
> 
> I'll also note that the version
> (e78ed7a67bbaa37dc57dfb37219f505c48755e14) is 3 years old. One big fix
> that happened since then is the change of some files' extensions for
> .C to .c, since they don't compile as .C .
> 
> As for why I'm interested in p9p, it's a feasible replacement base
> system in place of GNU--being small but still functional--, and sam
> and acme are useful editors.
> 
> On 6/30/17, Anthony G. Basile <basile@freeharbor.net> wrote:
>> On 6/30/17 3:02 AM, parazyd@dyne.org wrote:
>>> Do share them. plan9port doesn't compile on amd64 either.
>>>
>>> On Thu, 29 Jun 2017, Rendov Norra wrote:
>>>
>>>> I have patches to fix the first two issues (or at least they compile),
>>>> but they're very hacky as I'm not an experienced coder.
>>
>> Excellent, ignore my last message.  If you open a bug report at
>> bugs.gentoo.org, describe the problem and upload the patches and I'll
>> put it on the tree.  I also maintain plan9port so I can fix it directly.
>>
>> Having said that, I'm curious why people are interested in plan9port?
>>
>>>>
>>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>>> I should also note that this is on ARM
>>>>>
>>>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>>>> It fails at building its thread library, it fails at building
>>>>>> lib9/dirread, and I'm not sure if this is related at all but egrep
>>>>>> segfaults and kills the build.
>>>>>>
>>
>> --
>> Anthony G. Basile, Ph.D.
>> Gentoo Linux Developer [Hardened]
>> E-Mail    : basile@freeharbor.net
>> GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
>> GnuPG ID  : F52D4BBA
>>
>>
>>

I don't understand patch 2.  It eliminated an entire function which is
exposed for arm, so I don't see how that can be right.  Can you talk to
upstream about these.

-- 
Anthony G. Basile, Ph.D.
Gentoo Linux Developer [Hardened]
E-Mail    : blueness@gentoo.org
GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
GnuPG ID  : F52D4BBA

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[Rendov Norra]

I thought the functions commented out in the second patch are provided
by musl, but on further inspection, they are only declared in
ucontext.h, but are not implemented.

On 6/30/17, Anthony G. Basile <blueness@gentoo.org> wrote:
> On 6/30/17 5:14 PM, Rendov Norra wrote:
>> I've attached the patches here, but I hesitate to present them as a
>> solution since I'm incapable of testing them because of the segfaults
>> I'm experiencing with egrep.
>>
>> I'll also note that the version
>> (e78ed7a67bbaa37dc57dfb37219f505c48755e14) is 3 years old. One big fix
>> that happened since then is the change of some files' extensions for
>> .C to .c, since they don't compile as .C .
>>
>> As for why I'm interested in p9p, it's a feasible replacement base
>> system in place of GNU--being small but still functional--, and sam
>> and acme are useful editors.
>>
>> On 6/30/17, Anthony G. Basile <basile@freeharbor.net> wrote:
>>> On 6/30/17 3:02 AM, parazyd@dyne.org wrote:
>>>> Do share them. plan9port doesn't compile on amd64 either.
>>>>
>>>> On Thu, 29 Jun 2017, Rendov Norra wrote:
>>>>
>>>>> I have patches to fix the first two issues (or at least they compile),
>>>>> but they're very hacky as I'm not an experienced coder.
>>>
>>> Excellent, ignore my last message.  If you open a bug report at
>>> bugs.gentoo.org, describe the problem and upload the patches and I'll
>>> put it on the tree.  I also maintain plan9port so I can fix it directly.
>>>
>>> Having said that, I'm curious why people are interested in plan9port?
>>>
>>>>>
>>>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>>>> I should also note that this is on ARM
>>>>>>
>>>>>> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>>>>>>> It fails at building its thread library, it fails at building
>>>>>>> lib9/dirread, and I'm not sure if this is related at all but egrep
>>>>>>> segfaults and kills the build.
>>>>>>>
>>>
>>> --
>>> Anthony G. Basile, Ph.D.
>>> Gentoo Linux Developer [Hardened]
>>> E-Mail    : basile@freeharbor.net
>>> GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
>>> GnuPG ID  : F52D4BBA
>>>
>>>
>>>
>
> I don't understand patch 2.  It eliminated an entire function which is
> exposed for arm, so I don't see how that can be right.  Can you talk to
> upstream about these.
>
> --
> Anthony G. Basile, Ph.D.
> Gentoo Linux Developer [Hardened]
> E-Mail    : blueness@gentoo.org
> GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
> GnuPG ID  : F52D4BBA
>
>

Re: [gentoo-musl] Re: plan9port build fails on multiple levels.

[Anthony G. Basile]

On 6/29/17 9:01 PM, Rendov Norra wrote:
> I should also note that this is on ARM
>
> On 6/29/17, Rendov Norra <tsobf242@gmail.com> wrote:
>> It fails at building its thread library, it fails at building
>> lib9/dirread, and I'm not sure if this is related at all but egrep
>> segfaults and kills the build.
>>
You'll need to open a bug report outlining all the problems.  However, I
have to say, I don't think I have the time to really address plan9port
on musl, let along on arm.  Is there a good reason why this is useful to
people?  I ask because it really is a matter of the squeaky wheel gets
oiled first.

-- 
Anthony G. Basile, Ph.D.
Gentoo Linux Developer [Hardened]
E-Mail    : basile@freeharbor.net
GnuPG FP  : 1FED FAD9 D82C 52A5 3BAB  DC79 9384 FA6E F52D 4BBA
GnuPG ID  : F52D4BBA