freebsd-hackers Digest, Vol 365, Issue 5
freebsd-hac...@freebsd.org
freebsd...@freebsd.org
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
or, via email, send a message with subject or body 'help' to
freebsd-hac...@freebsd.org
You can reach the person managing the list at
freebsd-ha...@freebsd.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of freebsd-hackers digest..."
Today's Topics:
1. Re: Compiling kernel with gcc43 (Mario Lobo)
2. Re: Compiling kernel with gcc43 (Ryan Stone)
3. Re: Compiling kernel with gcc43 (Mario Lobo)
4. Re: Compiling kernel with gcc43 (Ryan Stone)
5. Re: Compiling kernel with gcc43 (Garrett Cooper)
6. Re: Compiling kernel with gcc43 (Andrius Mork?nas)
7. Re: Compiling kernel with gcc43 (Bruce Cran)
8. Re: Compiling kernel with gcc43 (Edward Tomasz Napiera?a)
9. Re: Compiling kernel with gcc43 [SOLVED] (Mario Lobo)
10. WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502 (Andriy Gapon)
11. Re: Another tool for updating /etc -- lua||other script
language bikeshed (per...@pluto.rain.com)
12. Re: Another tool for updating /etc -- lua||other script
language bikeshed (Robert Watson)
13. Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502 (Andriy Gapon)
14. Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502 (Bernhard Froehlich)
15. Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502 (Bernhard Froehlich)
16. Re: namei() returns EISDIR for "/" (Re: svn commit: r203990 -
head/lib/libc/sys) (Jaakko Heinonen)
17. Re: Another tool for updating /etc -- lua||other script
language bikeshed (Dag-Erling Sm?rgrav)
18. Re: namei() returns EISDIR for "/" (Re: svn commit: r203990 -
head/lib/libc/sys) (Alexander Best)
----------------------------------------------------------------------
Message: 1
Date: Thu, 25 Mar 2010 13:24:10 -0300
From: Mario Lobo <lo...@bsd.com.br>
Subject: Re: Compiling kernel with gcc43
To: FreeBSD-...@freebsd.org
Cc: freebsd...@freebsd.org
Message-ID:
<9884b5281003250924i462...@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
2010/3/24 Mario Lobo <lo...@bsd.com.br>
> Hi;
>
> I followed the instructions from
> http://www.freebsd.org/doc/en/articles/custom-gcc/article.html
>
> to see if I can achieve some performance gain (8-STABLE r205630 amd64) and
> also get a more modern instruction set for my Phenom II (amdfam10), so I've
> done the following config settings:
>
> make.conf
>
> .if !empty(.CURDIR:M/usr/ports/*) && exists(/usr/local/bin/gcc43)
> CC=/usr/local/bin/gcc43
> CXX=/usr/local/bin/g++43
> CPP=/usr/local/bin/cpp43
> CFLAGS+=-mssse3
> CPUTYPE=amdfam10
> .endif
>
> libmap.conf
>
> libgcc_s.so.1 gcc43/libgcc_s.so.1
> libgomp.so.1 gcc43/libgomp.so.1
> libobjc.so.3 gcc43/libobjc.so.2
> libssp.so.0 gcc43/libssp.so.0
> libstdc++.so.6 gcc43/libstdc++.so.6
>
>
> buildworld goes on normally it seems, but when I tried to build the kernel,
> the first stop came from the option used by the kernel build "-fformat-
> extensions", which is not accepted by gcc43, so I took it out of
> /usr/src/sys/conf/kern.mk, the ONLY place I found it.
>
> It seemed to solve the problem until gcc43 stopped because of the "-Werror"
> option, which is all over the place!. Funny that gcc42 (used by default)
> doesn't stop because of this option.
>
> Anyway,my question boils down to: Is there a way to build the kernel with
> gcc43, with minimal tweaking? Am i attempting something out of my league?
>
> Thanks,
> --
> Mario Lobo
>
>
Any suggestions/Comments?
Should I just dump the gcc43 idea and try this with clang/llvm?
I've seen http://wiki.freebsd.org/BuildingFreeBSDWithClang.
Anything else to watch out for when building world/kernel/ports?
Thanks,
--
Mario Lobo
http://www.mallavoodoo.com.br
FreeBSD since version 2.2.8 [not Pro-Audio.... YET!!] (99,7% winfoes FREE)
------------------------------
Message: 2
Date: Thu, 25 Mar 2010 13:04:17 -0400
From: Ryan Stone <rys...@gmail.com>
Subject: Re: Compiling kernel with gcc43
To: Mario Lobo <lo...@bsd.com.br>
Cc: freebsd...@freebsd.org, FreeBSD-...@freebsd.org
Message-ID:
<bc2d971003251004r737a...@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
If you just want to try it out you could disable -Werror for your own
build. The reason that gcc 4.3 fails where 4.2 works fine is that the
gcc developers are adding new warnings with every compiler release, so
until somebody takes the time to fix all of the new warnings you won't
be able to build the kernel with -Werror.
------------------------------
Message: 3
Date: Thu, 25 Mar 2010 14:18:34 -0300
From: Mario Lobo <lo...@bsd.com.br>
Subject: Re: Compiling kernel with gcc43
To: Ryan Stone <rys...@gmail.com>
Cc: freebsd...@freebsd.org, FreeBSD-...@freebsd.org
Message-ID:
<9884b5281003251018g259...@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
2010/3/25 Ryan Stone <rys...@gmail.com>
> If you just want to try it out you could disable -Werror for your own
> build. The reason that gcc 4.3 fails where 4.2 works fine is that the
> gcc developers are adding new warnings with every compiler release, so
> until somebody takes the time to fix all of the new warnings you won't
> be able to build the kernel with -Werror.
>
All right !! Thanks for replying !
There are a lot of locations throughout the source code where -Werror is
enabled
How can I disable -Werror globally? via src.conf ? will it do it for
world/kernel?
will this "damage" the resulting kernel/world binaries?
--
Mario Lobo
http://www.mallavoodoo.com.br
FreeBSD since version 2.2.8 [not Pro-Audio.... YET!!] (99,7% winfoes FREE)
------------------------------
Message: 4
Date: Thu, 25 Mar 2010 13:30:29 -0400
From: Ryan Stone <rys...@gmail.com>
Subject: Re: Compiling kernel with gcc43
To: Mario Lobo <lo...@bsd.com.br>
Cc: freebsd...@freebsd.org, FreeBSD-...@freebsd.org
Message-ID:
<bc2d971003251030u24c2...@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
-Wnoerror should disable it, but I don't know enough about the FreeBSD
build infrastructure to tell you where it needs to go. Worst comes to
worst you can just delete all of the -Werrors.
Disabling -Werror will not cause any problems. -Werror means "treat
warnings as errors". It's very useful for development but in your
case, where you're starting with a known good source tree with no
warnings with earlier compilers, it's not necessary to have the option
enabled. If the warnings are pointing to legitimate bugs, those bugs
also exist in a kernel built by gcc 4.2, so your kernel can't be any
worse than what you already have.
------------------------------
Message: 5
Date: Thu, 25 Mar 2010 10:31:07 -0700
From: Garrett Cooper <yane...@gmail.com>
Subject: Re: Compiling kernel with gcc43
To: Mario Lobo <lo...@bsd.com.br>
Cc: freebsd...@freebsd.org, Ryan Stone <rys...@gmail.com>,
FreeBSD-...@freebsd.org
Message-ID:
<7d6fde3d1003251031v35...@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
On Thu, Mar 25, 2010 at 10:18 AM, Mario Lobo <lo...@bsd.com.br> wrote:
> 2010/3/25 Ryan Stone <rys...@gmail.com>
>
>> If you just want to try it out you could disable -Werror for your own
>> build. The reason that gcc 4.3 fails where 4.2 works fine is that the
>> gcc developers are adding new warnings with every compiler release, so
>> until somebody takes the time to fix all of the new warnings you won't
>> be able to build the kernel with -Werror.
>>
>
> All right !! Thanks for replying !
>
> There are a lot of locations throughout the source code where -Werror is
> enabled
> How can I disable -Werror globally? via src.conf ? will it do it for
> world/kernel?
> will this "damage" the resulting kernel/world binaries?
-Wno-error .
-Garrett
------------------------------
Message: 6
Date: Thu, 25 Mar 2010 19:12:42 +0200
From: Andrius Mork?nas <hino...@gmail.com>
Subject: Re: Compiling kernel with gcc43
To: "Mario Lobo" <lo...@bsd.com.br>, FreeBSD-...@freebsd.org
Cc: freebsd...@freebsd.org
Message-ID: <op.u94vffbk43o42p@klevas>
Content-Type: text/plain; charset=utf-8; format=flowed; delsp=yes
On Thu, 25 Mar 2010 18:24:10 +0200, Mario Lobo <lo...@bsd.com.br> wrote:
> Should I just dump the gcc43 idea and try this with clang/llvm?
>
> I've seen http://wiki.freebsd.org/BuildingFreeBSDWithClang.
>
> Anything else to watch out for when building world/kernel/ports?
ClangBSD had little to no runtime testing, so you should be careful
with it. I'm not saying it won't work, but make sure you have backups
in case something goes wrong.
Also, if you just want to build the kernel, you don't need ClangBSD,
you can just install llvm-devel from ports and use
> make CC=/usr/local/bin/clang buildkernel
Some people ran or are running their systems with clang compiled kernel.
You should still be careful, and some things (like snd_hda on amd64)
are known to have problems.
--
Andrius
------------------------------
Message: 7
Date: Thu, 25 Mar 2010 18:30:21 +0000
From: Bruce Cran <bru...@muon.cran.org.uk>
Subject: Re: Compiling kernel with gcc43
To: Mario Lobo <lo...@bsd.com.br>
Cc: freebsd...@freebsd.org, Ryan Stone <rys...@gmail.com>,
FreeBSD-...@freebsd.org
Message-ID: <20100325183...@muon.cran.org.uk>
Content-Type: text/plain; charset=us-ascii
On Thu, Mar 25, 2010 at 02:18:34PM -0300, Mario Lobo wrote:
> All right !! Thanks for replying !
>
> There are a lot of locations throughout the source code where -Werror is
> enabled
> How can I disable -Werror globally? via src.conf ? will it do it for
> world/kernel?
> will this "damage" the resulting kernel/world binaries?
You should be able to use NO_WERROR in src.conf to prevent -Werror
being used.
--
Bruce Cran
------------------------------
Message: 8
Date: Thu, 25 Mar 2010 20:12:09 +0100
From: Edward Tomasz Napiera?a <tr...@FreeBSD.org>
Subject: Re: Compiling kernel with gcc43
To: Mario Lobo <lo...@bsd.com.br>
Cc: freebsd...@freebsd.org, FreeBSD-...@freebsd.org
Message-ID: <52C0E835-3C10-47B6...@FreeBSD.org>
Content-Type: text/plain; charset=iso-8859-2
Wiadomość napisana przez Mario Lobo w dniu 2010-03-25, o godz. 17:24:
> 2010/3/24 Mario Lobo <lo...@bsd.com.br>
>>
>> Anyway,my question boils down to: Is there a way to build the kernel with
>> gcc43, with minimal tweaking? Am i attempting something out of my league?
This is what I use to compile kernel on 9.0-CURRENT/amd64:
CC= /usr/local/bin/gcc44
CXX= /usr/local/bin/g++44
CWARNFLAGS= -Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes \
-Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual \
-Wno-pointer-sign -D__FreeBSD_cc_version=0 \
-Wno-unused-function -Wno-inline -Wno-format \
-Wno-uninitialized -Wno-array-bounds -Wno-missing-prototypes
Note that this is GCC 4.4.4, installed from ports. Also, this will not
compile the world - only the kernel. Also, they were a few fixes to make
this work; they were not MFC-ed, iirc; removing -Werror altogether should
work around these problems.
--
If you cut off my head, what would I say? Me and my head, or me and my body?
------------------------------
Message: 9
Date: Thu, 25 Mar 2010 18:43:36 +0000
From: Mario Lobo <lo...@bsd.com.br>
Subject: Re: Compiling kernel with gcc43 [SOLVED]
To: FreeBSD-...@freebsd.org
Cc: freebsd...@freebsd.org
Message-ID: <20100325184...@bsd.com.br>
Content-Type: Text/Plain; charset="iso-8859-2"
On Thursday 25 March 2010 19:12:09 Edward Tomasz Napierała wrote:
> Wiadomość napisana przez Mario Lobo w dniu 2010-03-25, o godz. 17:24:
> > 2010/3/24 Mario Lobo <lo...@bsd.com.br>
> >
> >> Anyway,my question boils down to: Is there a way to build the kernel
> >> with gcc43, with minimal tweaking? Am i attempting something out of my
> >> league?
>
> This is what I use to compile kernel on 9.0-CURRENT/amd64:
>
> CC= /usr/local/bin/gcc44
> CXX= /usr/local/bin/g++44
> CWARNFLAGS= -Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes
> \ -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual \
> -Wno-pointer-sign -D__FreeBSD_cc_version=0 \
> -Wno-unused-function -Wno-inline -Wno-format \
> -Wno-uninitialized -Wno-array-bounds -Wno-missing-prototypes
>
> Note that this is GCC 4.4.4, installed from ports. Also, this will not
> compile the world - only the kernel. Also, they were a few fixes to make
> this work; they were not MFC-ed, iirc; removing -Werror altogether should
> work around these problems.
>
> --
> If you cut off my head, what would I say? Me and my head, or me and my
> body?
>
Adding
NO_WERROR=
WERROR=
to src.conf did it. It's compiling right now.
I'll post my findings and impressions on results and performance right after
the next reboot.
Thanks for all the help and info, Guys !!
--
Mario Lobo
http://www.mallavoodoo.com.br
FreeBSD since version 2.2.8 [not Pro-Audio.... YET!!] (99,7% winfoes FREE)
------------------------------
Message: 10
Date: Fri, 26 Mar 2010 01:41:21 +0200
From: Andriy Gapon <a...@icyb.net.ua>
Subject: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502
To: freebsd-...@freebsd.org, vb...@freebsd.org,
freebsd...@freebsd.org
Message-ID: <4BABF4A1...@icyb.net.ua>
Content-Type: text/plain; charset=UTF-8
On VirtualBox startup the following messages are produced in system log:
WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl ffffffffc4a81502
This happens on amd64 platform.
I think the reason for this is in RTFileIoCtl() helper function (in
VBox/Runtime/r3/posix/fileio-posix.cpp).
This function takes iRequest argument as 'int' and then passes it as the second
parameter to ioctl(2). But on FreeBSD that parameter, request, is expected to
be of type unsigned long.
Thus, a sufficiently large value of a request gets represented as a negative
value in iRequest, which then leads to the warning.
On the other hand, looking at how ioctl(2) is declared in other operating
systems, I am not sure who is right and who is wrong here.
Maybe we should a cast to unsigned int in RTFileIoCtl() for __FreeBSD__ case.
Maybe FreeBSD should just suck it up and remove the noisy warning.
I know too little to suggest a resolution.
P.S. it seems that Mac OS X is in the same boat as FreeBSD, perhaps other BSDs too.
--
Andriy Gapon
------------------------------
Message: 11
Date: Fri, 26 Mar 2010 00:49:10 -0700
From: per...@pluto.rain.com
Subject: Re: Another tool for updating /etc -- lua||other script
language bikeshed
To: rwa...@freebsd.org
Cc: freebsd...@freebsd.org
Message-ID: <4bac66f6.4sYPjsI4heo4SpWm%per...@pluto.rain.com>
Content-Type: text/plain; charset=us-ascii
Robert Watson <rwa...@freebsd.org> wrote:
> ... web browsers [are] basically operating systems at this point ...
Isn't this a bit of an exaggeration? Not too many browsers have
to deal with process/thread scheduling, or device drivers, or
booting, or file system issues -- they rely on the OS for that
(as does any other application).
------------------------------
Message: 12
Date: Fri, 26 Mar 2010 08:45:10 +0000 (GMT)
From: Robert Watson <rwa...@FreeBSD.org>
Subject: Re: Another tool for updating /etc -- lua||other script
language bikeshed
To: per...@pluto.rain.com
Cc: freebsd...@freebsd.org
Message-ID: <alpine.BSF.2.00.1...@fledge.watson.org>
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
On Fri, 26 Mar 2010, per...@pluto.rain.com wrote:
> Robert Watson <rwa...@freebsd.org> wrote:
>> ... web browsers [are] basically operating systems at this point ...
>
> Isn't this a bit of an exaggeration? Not too many browsers have to deal
> with process/thread scheduling, or device drivers, or booting, or file
> system issues -- they rely on the OS for that (as does any other
> application).
I think it's more of an anaology than an exageration. The FreeBSD kernel,
including device drivers and architectures, is around 3.9 million lines of
code. Google's Chromium, including WebKit, is around 4.1 million lines of
code. Both provide an extensive runtime environment for applications that run
on top of them, security domains, storage services, and management models.
I'm not arguing that web browsers are a substitute for our current operating
system layer: they clearly build on it. However, in terms of their goals in
providing an execution environment, user interface, etc, they fill a very
similar niche by being a general-purpose platform for many specific things.
And, to get back to the point I was making: if you toast your Chromium update
or get configuration management wrong, then your applications (Google Docs,
GMail, ...) on ChromeOS won't work any more than if you toasted your /lib or
/etc in FreeBSD. For example, if the Chromium configuration files change and
it forgets about web proxies, Chromium won't be able to call home to pick up a
fix any more than if etcmerge toasts resolv.conf.
Making updates easy is, to a large extent, about avoiding the creation of
foot-shooting opportunities. Some of it is about tools (binary updates,
mergers, rollbac, etc), but most of it is about avoiding scenarios in which a
previously valid configuration becomes invalid. And if we look at problems
FreeBSD has had with updates in a past, a lot come down to precisely that: for
example, renaming serial port device names (several times in as many years).
Robert
------------------------------
Message: 13
Date: Fri, 26 Mar 2010 11:14:49 +0200
From: Andriy Gapon <a...@icyb.net.ua>
Subject: Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502
To: Bernhard Froehlich <de...@bluelife.at>
Cc: freebsd...@freebsd.org, freebsd-...@freebsd.org,
vb...@freebsd.org
Message-ID: <4BAC7B09...@icyb.net.ua>
Content-Type: text/plain; charset=UTF-8
on 26/03/2010 10:59 Bernhard Froehlich said the following:
>
> I've had a talk to Alexander Eichner and they have already fixed that 2
> months ago in r25953 on trunk. So VirtualBox 3.2.x will include this fix.
>
> http://www.virtualbox.org/changeset/25953
Thank you very much.
Sorry for not checking the upstream repository first.
--
Andriy Gapon
------------------------------
Message: 14
Date: Fri, 26 Mar 2010 10:18:57 +0100
From: Bernhard Froehlich <de...@bluelife.at>
Subject: Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502
To: Andriy Gapon <a...@icyb.net.ua>
Cc: freebsd...@freebsd.org, freebsd-...@freebsd.org,
vb...@freebsd.org
Message-ID: <6cdec968b118db90...@bluelife.at>
Content-Type: text/plain; charset=UTF-8
On Fri, 26 Mar 2010 11:14:49 +0200, Andriy Gapon <a...@icyb.net.ua> wrote:
> on 26/03/2010 10:59 Bernhard Froehlich said the following:
>>
>> I've had a talk to Alexander Eichner and they have already fixed that 2
>> months ago in r25953 on trunk. So VirtualBox 3.2.x will include this
>> fix.
>>
>> http://www.virtualbox.org/changeset/25953
>
> Thank you very much.
> Sorry for not checking the upstream repository first.
No problem. We will backport the fix and include it with the port update
to 3.1.6. beat is currently testing it.
--
Bernhard FrĂśhlich
http://www.bluelife.at/
------------------------------
Message: 15
Date: Fri, 26 Mar 2010 09:59:34 +0100
From: Bernhard Froehlich <de...@bluelife.at>
Subject: Re: WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl
ffffffffc4a81502
To: Andriy Gapon <a...@icyb.net.ua>
Cc: freebsd...@freebsd.org, freebsd-...@freebsd.org,
vb...@freebsd.org
Message-ID: <0195a63b519a4457...@bluelife.at>
Content-Type: text/plain; charset=UTF-8
On Fri, 26 Mar 2010 01:41:21 +0200, Andriy Gapon <a...@icyb.net.ua> wrote:
> On VirtualBox startup the following messages are produced in system log:
> WARNING pid ... (VBoxSVC): ioctl sign-extension ioctl ffffffffc4a81502
>
> This happens on amd64 platform.
>
> I think the reason for this is in RTFileIoCtl() helper function (in
> VBox/Runtime/r3/posix/fileio-posix.cpp).
> This function takes iRequest argument as 'int' and then passes it as the
> second
> parameter to ioctl(2). But on FreeBSD that parameter, request, is
> expected to
> be of type unsigned long.
> Thus, a sufficiently large value of a request gets represented as a
> negative
> value in iRequest, which then leads to the warning.
>
> On the other hand, looking at how ioctl(2) is declared in other
operating
> systems, I am not sure who is right and who is wrong here.
> Maybe we should a cast to unsigned int in RTFileIoCtl() for __FreeBSD__
> case.
> Maybe FreeBSD should just suck it up and remove the noisy warning.
>
> I know too little to suggest a resolution.
>
> P.S. it seems that Mac OS X is in the same boat as FreeBSD, perhaps
other
> BSDs too.
I've had a talk to Alexander Eichner and they have already fixed that 2
months ago in r25953 on trunk. So VirtualBox 3.2.x will include this fix.
http://www.virtualbox.org/changeset/25953
--
Bernhard FrĂśhlich
http://www.bluelife.at/
------------------------------
Message: 16
Date: Fri, 26 Mar 2010 13:41:00 +0200
From: Jaakko Heinonen <j...@FreeBSD.org>
Subject: Re: namei() returns EISDIR for "/" (Re: svn commit: r203990 -
head/lib/libc/sys)
To: freebsd...@FreeBSD.org
Cc: Alexander Best <alexb...@wwu.de>, Poul-Henning Kamp
<p...@phk.freebsd.dk>, Bruce Evans <br...@optusnet.com.au>
Message-ID:
<2010032611...@a91-153-117-195.elisa-laajakaista.fi>
Content-Type: text/plain; charset=us-ascii
On 2010-03-05, Jaakko Heinonen wrote:
> I have updated the patch taking some of bde's comments into account. The
> new version also includes updates for namei(9) manual page.
Yet another update:
http://people.freebsd.org/~jh/patches/lookup-root.4.diff
I have committed the relookup() part as r205682.
Unfortunately errno translation is needed also for open("/", O_CREAT)
and undelete("/").
--
Jaakko
------------------------------
Message: 17
Date: Fri, 26 Mar 2010 12:46:17 +0100
From: Dag-Erling Sm?rgrav <d...@des.no>
Subject: Re: Another tool for updating /etc -- lua||other script
language bikeshed
To: Ivan Voras <ivo...@freebsd.org>
Cc: freebsd...@freebsd.org
Message-ID: <86pr2rm...@ds4.des.no>
Content-Type: text/plain; charset=utf-8
Ivan Voras <ivo...@freebsd.org> writes:
> Dag-Erling Smørgrav <d...@des.no> writes:
> > Bourne shell is a perfectly fine programming language if you know
> > how to use it.
> I'll agree that it's fine but only in the abstract - e.g. that it is
> Turing complete :)
Emphasis on "if you know how to use it". It's like Perl - it's easy to
learn how to write bad Perl or Bourne shell code, but much harder to
learn how to use them properly.
DES
--
Dag-Erling Smørgrav - d...@des.no
------------------------------
Message: 18
Date: Fri, 26 Mar 2010 12:56:58 +0100 (CET)
From: Alexander Best <alexb...@wwu.de>
Subject: Re: namei() returns EISDIR for "/" (Re: svn commit: r203990 -
head/lib/libc/sys)
To: Jaakko Heinonen <j...@FreeBSD.org>, <freebsd...@FreeBSD.org>
Cc: Poul-Henning Kamp <p...@phk.freebsd.dk>, Bruce Evans
<br...@optusnet.com.au>
Message-ID:
<permail-20100326115658f...@message-id.uni-muenster.de>
Content-Type: text/plain; charset=us-ascii
Jaakko Heinonen schrieb am 2010-03-26:
> On 2010-03-05, Jaakko Heinonen wrote:
> > I have updated the patch taking some of bde's comments into
> > account. The
> > new version also includes updates for namei(9) manual page.
> Yet another update:
> http://people.freebsd.org/~jh/patches/lookup-root.4.diff
> I have committed the relookup() part as r205682.
> Unfortunately errno translation is needed also for open("/", O_CREAT)
> and undelete("/").
you might want to consider posting your patch on
http://reviews.freebsdish.org. that way it'll be easier to review it.
--
Alexander Best
------------------------------
End of freebsd-hackers Digest, Vol 365, Issue 5
***********************************************