DragonFlyBSD bugtracker: Issueshttps://bugs.dragonflybsd.org/https://bugs.dragonflybsd.org/favicon.ico?16293952082011-12-11T00:45:43ZDragonFlyBSD bugtracker
Redmine DragonFlyBSD - Submit #2256 (Closed): [PATCH] Added the variable NO_GCC44 to the systemhttps://bugs.dragonflybsd.org/issues/22562011-12-11T00:45:43Zjuanfra684
<p>The purpose of this change is allow to the developers skip the compilation of<br />GCC 4.1 and 4.4 and save the time spent in the compilation of GCC, when they<br />are compiling the system with other compiler (like clang).<br />---<br /> Makefile.inc1 | 4 +<ins>++<br /> etc/defaults/make.conf | 1 +<br /> gnu/lib/Makefile | 5 +</ins>++-<br /> gnu/usr.bin/Makefile | 2 +<ins><br /> share/man/man5/make.conf.5 | 3 +</ins>+<br /> 5 files changed, 14 insertions(+), 1 deletions(-)</p> DragonFlyBSD - Bug #2253 (Closed): Weird behavior of some moduleshttps://bugs.dragonflybsd.org/issues/22532011-12-06T00:37:14Zjuanfra684
<p>I've had this bug in my notes from one or two months. I report this now because I think is related with <a class="issue tracker-1 status-1 priority-3 priority-lowest" title="Bug: snd_hda not useable if loaded via /boot/loader.conf (New)" href="https://bugs.dragonflybsd.org/issues/2252">#2252</a> but the problem is more wide than the specific modules. My problem is with dsched_fq.</p>
<p>Steps: <br /> - Delete or comment all lines related to dsched in loader.conf and sysctl.conf.<br /> - Reboot the system.<br /> - Log in the system with the root account.<br /> - Run "sysctl dsched.policies":
# dsched.policies = noop fq<br /> - Run "kldstat". It doesn't show the module dsched_fq loaded.<br /> - Run "sysctl -w dsched.policy.default=fq":
# noop -> fq</p>
<p>All works perfect, right?. More steps:<br /> - Add dsched_fq to your loader.conf file.<br /> - Reboot the system.<br /> - Log in the system with the root account.<br /> - Run "sysctl dsched.policies":
# dsched.policies = noop<br /> - Run "kldstat". It shows the module dsched_fq loaded.<br /> - Run "sysctl -w dsched.policy.default=fq":
# noop -> noop</p>
<p>I updated the past week the base and kernel (i386)</p> DragonFlyBSD - Bug #2245 (New): panic: assertion "ref < &td->td_toks_end" failed in lwkt_gettoken...https://bugs.dragonflybsd.org/issues/22452011-11-23T03:00:26Zjuanfra684
<p>I don't know if this is important or not, so you can close the bug if the info isn't useful. Don't worry. I don't need a fix. I've opened the bug only for share the core.txt.</p>
<p>I was compiling a package in the console (the real console, no in a X terminal) and I closed my Xorg session with [Ctrl]+[C]. The system crashed.</p>
<p><a class="external" href="http://juanfra.info/bugs-y-listas/crash7-201111/core.txt.7.bz2">http://juanfra.info/bugs-y-listas/crash7-201111/core.txt.7.bz2</a></p> DragonFlyBSD - Submit #2208 (Closed): [PATCH] Enhancement the support for clang in compilers.confhttps://bugs.dragonflybsd.org/issues/22082011-11-17T02:40:45Zjuanfra684
<p>- It should not select the standards.</p>
<pre><code>- Commented the lines related to clang++. It breaks buildworld.</code></pre>
<pre><code>- VERSION changed to 4.4.7 for avoid software compiling with g++ 4.1.</code></pre>
<pre><code>Authors: Sascha Wildner &lt;<a class="email" href="mailto:saw@online.de">saw@online.de</a>&gt;<br /> Juan Francisco Cantero Hurtado &lt;<a class="email" href="mailto:iam@juanfra.info">iam@juanfra.info</a>&gt;<br />---<br /> etc/defaults/compilers.conf | 18 +<ins>++</ins>+++++++------<br /> 1 files changed, 12 insertions(+), 6 deletions(-)</code></pre>
<p>diff --git a/etc/defaults/compilers.conf b/etc/defaults/compilers.conf<br />index 9952679..3e7526c 100644<br />--- a/etc/defaults/compilers.conf<br />+++ b/etc/defaults/compilers.conf<br /><code>@ -1,16 +1,22 </code>@
# clang from SVN trunk
#<br /> clangsvn_CC=/usr/local/bin/clang<br />-clangsvn_CXX=/usr/local/bin/clang++<br />-clangsvn_CFLAGS="-fno-color-diagnostics -Qunused-arguments -std=gnu89" <br />-clangsvn_CXXFLAGS="-fno-color-diagnostics -Qunused-arguments -std=gnu++98" <br />-clangsvn_VERSION=4.1.2<br /><ins>clangsvn_CFLAGS="-fno-color-diagnostics -Qunused-arguments" <br /></ins># Uncomment this lines for compile C++ code with clang.<br /><ins># Disabled by default because clang</ins>+ breaks buildworld.<br /><ins>#clangsvn_CXX=/usr/local/bin/clang</ins>+<br />+#clangsvn_CXXFLAGS="-fno-color-diagnostics -Qunused-arguments" <br />+clangsvn_VERSION=4.4.7</p>
<ol>
<li>clang from pkgsrc
#<br /> clang_CC=/usr/pkg/bin/clang<br />-clang_CFLAGS="-fno-color-diagnostics -Qunused-arguments -std=gnu89" <br />-clang_VERSION=4.1.2<br /><ins>clang_CFLAGS="-fno-color-diagnostics -Qunused-arguments" <br /></ins># Uncomment this lines for compile C++ code with clang.<br /><ins># Disabled by default because clang</ins>+ breaks buildworld.<br /><ins>#clang_CXX=/usr/pkg/bin/clang</ins>+<br />+#clang_CXXFLAGS="-fno-color-diagnostics -Qunused-arguments" <br />+clang_VERSION=4.4.7</li>
</ol>
<ol>
<li>gnat-aux (i.e. gcc 4.6) from pkgsrc
#<br />-- <br />Juan Francisco Cantero Hurtado <a class="external" href="http://juanfra.info">http://juanfra.info</a></li>
</ol> DragonFlyBSD - Submit #2206 (Closed): [PATCH] Libmagic doesn't compile with clanghttps://bugs.dragonflybsd.org/issues/22062011-11-17T02:20:53Zjuanfra684
<ul>
<li><del>I</del> is obsolete. Changed to -iquote.
<ul>
<li>gcc is ignoring file.h. Fix for the error compiling with clang.<br />---<br /> lib/libmagic/Makefile | 6 +<ins>+---<br /> 1 files changed, 3 insertions(</ins>), 3 deletions(-)</li>
</ul></li>
</ul>
<p>diff --git a/lib/libmagic/Makefile b/lib/libmagic/Makefile<br />index e339983..2151523 100644<br />--- a/lib/libmagic/Makefile<br />+<ins>+ b/lib/libmagic/Makefile<br /><code>@ -54,9 +54,9 </code>@ magic.mgc: mkmagic.nx magic<br /> CLEANFILES</ins>= mkmagic.nx<br /> build-tools: mkmagic.nx</p>
<p><del>NXCFLAGS= ${LIBMAGIC_CDEFS} -DCOMPILE_ONLY -I${.OBJDIR} -I</del> <del>I${.CURDIR} -I${CONTRIBDIR}<br /><ins>NXCFLAGS= ${LIBMAGIC_CDEFS} -DCOMPILE_ONLY -I${.OBJDIR} -iquote -I${.CURDIR} -I${CONTRIBDIR}<br /> NXCFLAGS</ins>= -DBOOTSTRAPPING<br />-mkmagic.nx: file.h apprentice.c funcs.c magic.c print.c<br /></del> ${NXCC} ${NXCFLAGS} ${NXLDFLAGS} ${.ALLSRC} ${NXLDLIBS} -o ${.TARGET}<br /><ins>mkmagic.nx: apprentice.c funcs.c magic.c print.c<br /></ins> ${NXCC} ${NXCFLAGS} ${NXLDFLAGS} ${.ALLSRC} ${NXLDLIBS} -include ${CONTRIBDIR}/file.h -o ${.TARGET}</p>
<pre><code>.include &lt;bsd.lib.mk&gt;<br />-- <br />Juan Francisco Cantero Hurtado <a class="external" href="http://juanfra.info">http://juanfra.info</a></code></pre> DragonFlyBSD - Submit #2205 (Closed): [PATCH] Fixes in libexec/custom/cchttps://bugs.dragonflybsd.org/issues/22052011-11-17T02:20:51Zjuanfra684
<p>- Removed unnecessary code with the actual versions of clang.</p>
<pre><code>- Removed the conditional code for cpp.</code></pre>
<pre><code>Authors: Sascha Wildner &lt;<a class="email" href="mailto:saw@online.de">saw@online.de</a>&gt;<br /> Juan Francisco Cantero Hurtado &lt;<a class="email" href="mailto:iam@juanfra.info">iam@juanfra.info</a>&gt;<br />---<br /> libexec/customcc/cc.sh | 28 +<ins>+-------------------------<br /> 1 files changed, 3 insertions(</ins>), 25 deletions(-)</code></pre>
<p>diff --git a/libexec/customcc/cc.sh b/libexec/customcc/cc.sh<br />index 8755ef6..ae537af 100644<br />--- a/libexec/customcc/cc.sh<br />+++ b/libexec/customcc/cc.sh<br /><code>@ -3,31 +3,9 </code>@<br /> CDIR=$(dirname $0)<br /> CNAME=$(basename $0)</p>
<p><del># XXX clang needs some special handling<br /></del>#<br /><del># it is called only for "cc" and "gcc" and even then it could have been<br /></del># run on c++ files<br /><del>#<br />-if [ "${CCVER}" = "clang" ]; then<br /></del> if [ "${CNAME}" = "cpp" ]; then<br />- exec ${CDIR}/../gcc41/cpp "$<code>" <br />- elif [ "${CNAME}" = "c++" -o "${CNAME}" = "g++" ]; then<br />- exec ${CDIR}/../gcc41/c++ "$</code>" <br />- elif [ <del>z $beenhere ]; then<br /></del> export beenhere=1<br />- oldargs="$<code>" <br />- export oldargs<br />- INCOPT="-nobuiltininc -nostdinc \<br />- -isysroot </code>@INCPREFIX@<code> \<br />- -isystem /usr/include \<br />- -isystem /usr/libdata/gcc41 \<br />- -isystem /usr/include/c++/4.1" <br />- elif [ "${CNAME}" = "cc" -o "${CNAME}" = "gcc" ]; then<br />- exec ${CDIR}/../gcc41/cc $oldargs<br />- fi<br />-elif [ "${CCVER}" = "clangsvn" ]; then<br />- if [ "${CNAME}" = "cpp" ]; then<br />- exec ${CDIR}/../gcc41/cpp "$</code>" <br /><ins>if [ "${CCVER}" = "clang" <del>o "${CCVER}" = "clangsvn" ]; then<br /></ins> if [ "${CNAME}" = "g++" -o "${CNAME}" = "c++" -o "${CNAME}" = "cc1plus" ]; then<br />+ exec ${CDIR}/../gcc44/g++ "$<code>" <br /> else<br /> INCOPT="-nobuiltininc -nostdinc \<br /> -isysroot </code>@INCPREFIX@@ \<br />-</del> <br />Juan Francisco Cantero Hurtado <a class="external" href="http://juanfra.info">http://juanfra.info</a></p> DragonFlyBSD - Bug #2179 (Closed): The libc in dfly/amd64 hasn't "getcontext"https://bugs.dragonflybsd.org/issues/21792011-11-01T10:21:47Zjuanfra684
<p>Reported by John Marino in<br /><a class="external" href="http://gnats.netbsd.org/cgi-bin/query-pr-single.pl?number=45500">http://gnats.netbsd.org/cgi-bin/query-pr-single.pl?number=45500</a>.</p>
<p>The version i386 of libc has "getcontext" but the amd64 version not.<br />The work in "getcontext" is previous to GSoC/amd64 (one and half year). I think<br />the GSoC project didn't ported "getcontext" to amd64.</p>
<p>I've set the priority to bug, not feature request, because this is a<br />inconsistency libc/i386 vs libc/amd64.</p>
<p>References:<br />- lib/libc/i386/sys/getcontext.S<br />- Request for swapcontext and getcontext to be ported to our libc:<br /><a class="external" href="http://leaf.dragonflybsd.org/mailarchive/users/2006-12/msg00153.html">http://leaf.dragonflybsd.org/mailarchive/users/2006-12/msg00153.html</a><br />- Port DragonFly to the AMD64 architecture:<br /><a class="external" href="http://www.dragonflybsd.org/docs/developer/AMD64GSoC/">http://www.dragonflybsd.org/docs/developer/AMD64GSoC/</a></p> DragonFlyBSD - Submit #2178 (Closed): [PATCH] Note about the use of secondary cards with DRIhttps://bugs.dragonflybsd.org/issues/21782011-11-01T09:54:14Zjuanfra684
<p>---<br /> share/man/man4/drm.4 | 6 +<ins>++</ins>+<br /> 1 files changed, 6 insertions(+), 0 deletions(-)</p>
<p>diff --git a/share/man/man4/drm.4 b/share/man/man4/drm.4<br />index 3c38f08..688e388 100644<br />--- a/share/man/man4/drm.4<br />+<ins>+ b/share/man/man4/drm.4<br /><code>@ -102,6 +102,12 </code>@ EndSection<br /> .Ed<br /> .El<br /> .Pp<br />+If you have two graphic cards (i.e.: a on-board graphics and other AGP/PCI/PCIe<br />+card) and you would like use the secondary device with DRI, you will need make<br />+a symbolic link from /dev/dri/card1 to /dev/dri/card0 always. This is a old bug<br />+of Xorg, the DRI module only reads the first device.<br /></ins>.El<br />+.Pp<br /> Debugging output can be enabled and disabled by setting the<br /> .Xr sysctl 8<br /> node<br />-- <br />1.7.7</p> DragonFlyBSD - Bug #2173 (Resolved): The installer should not add usbd to rc.confhttps://bugs.dragonflybsd.org/issues/21732011-10-31T08:43:25Zjuanfra684
<p>The installer adds by default "usbd" to "rc.conf". With "usbd" running, Xorg<br />can't use my usb mouse (ums0) because the device is busy. With<br />"usbd"+"moused"+"/dev/sysmouse" in xorg.conf, the mouse doesn't work.</p>
<p>So, the installer installs a system with mouse support in console (is it<br />useful?) but blocking the correct operation of the mouse in X.</p> DragonFlyBSD - Bug #2160 (Closed): [PATCH 2/2] New option in adduser for to set the mode of homehttps://bugs.dragonflybsd.org/issues/21602011-10-26T20:11:53Zjuanfra684
<p>Imported the last version of adduser from FreeBSD 9.0-CURRENT. The only<br />change is the new <del>M option.<br />--</del><br /> usr.sbin/adduser/adduser.8 | 4 +<ins>++<br /> usr.sbin/adduser/adduser.sh | 40 +</ins>++++++++++++++++++++++++++++++++++++++<br /> 2 files changed, 44 insertions(+), 0 deletions(-)</p>
<p>diff --git a/usr.sbin/adduser/adduser.8 b/usr.sbin/adduser/adduser.8<br />index aad06be..fbdfe75 100644<br />--- a/usr.sbin/adduser/adduser.8<br />+<ins>+ b/usr.sbin/adduser/adduser.8<br /><code>@ -38,6 +38,7 </code>@<br /> .Op Fl CDENShq<br /> .Op Fl G Ar groups<br /> .Op Fl L Ar login_class<br /></ins>.Op Fl M Ar mode<br /> .Op Fl d Ar partition<br /> .Op Fl f Ar file<br /> .Op Fl g Ar login_group<br /><code>@ -240,6 +241,9 </code>@ Please note that the message<br /> file can reference the internal variables of the<br /> .Nm<br /> script.<br /><ins>.It Fl M Ar mode<br />+Create the home directory with permissions set to<br /></ins>.Ar mode .<br /> .It Fl N<br /> Do not read the default configuration file.<br /> .It Fl q<br />diff --git a/usr.sbin/adduser/adduser.sh b/usr.sbin/adduser/adduser.sh<br />index 3a8a9d3..500dbb0 100644<br />--- a/usr.sbin/adduser/adduser.sh<br />+<ins>+ b/usr.sbin/adduser/adduser.sh<br /><code>@ -82,6 +82,7 </code>@ show_usage() {<br /> echo " -E disable this account after creation" <br /> echo " -G additional groups to add accounts to" <br /> echo " -L login class of the user" <br /></ins> echo " -M file permission for home directory" <br /> echo " -N do not read configuration file" <br /> echo " -S a nonexistent shell is not an error" <br /> echo " -d home directory" <br /><code>@ -191,6 +192,7 </code>@ save_config() {<br /> echo "# NOTE: only <strong>some</strong> variables are saved." >> ${ADDUSERCONF}<br /> echo "# Last Modified on `${DATECMD}`." >> ${ADDUSERCONF}<br /> echo '' >> ${ADDUSERCONF}<br />+ echo "defaultHomePerm=$uhomeperm" >> ${ADDUSERCONF}<br /> echo "defaultLgroup=$ulogingroup" >> ${ADDUSERCONF}<br /> echo "defaultclass=$uclass" >> ${ADDUSERCONF}<br /> echo "defaultgroups=$ugroups" >> ${ADDUSERCONF}<br /><code>@ -248,6 +250,12 </code>@ add_user() {<br /> if [ "$uhome" = "$NOHOME" ]; then<br /> _home='-d "$uhome"'<br /> else<br />+ # Use home directory permissions if specified<br />+ if [ -n "$uhomeperm" ]; then<br />+ _home='-m -d "$uhome" -M "$uhomeperm"'<br />+ else<br />+ _home='-m -d "$uhome"'<br />+ fi<br /> _home='-m -d "$uhome"'<br /> fi<br /> elif [ -n "$Dflag" -a -n "$uhome" ]; then<br /><code>@ -450,6 +458,29 </code>@ get_homedir() {<br /> fi<br /> }</p>
<p><ins># get_homeperm<br /></ins># Reads the account's home directory permissions<br /><ins>#<br />+get_homeperm() {<br /></ins> uhomeperm=$defaultHomePerm<br />+ _input=<br />+ _prompt=<br /><ins><br /></ins> if [ -n "$uhomeperm" ]; then<br />+ _prompt="Home directory permissions [${uhomeperm}]: " <br />+ else<br />+ _prompt="Home directory permissions (Leave empty for default): " <br />+ fi<br />+ if [ -z "$fflag" ]; then<br />+ echo -n "$_prompt" <br />+ read _input<br />+ fi<br /><ins><br /></ins> if [ -n "$_input" ]; then<br />+ uhomeperm="$_input" <br />+ fi<br /><ins>}<br /></ins>
# get_uid
# Reads a numeric userid in an interactive or batch session. Automatically
# allocates one if it is not specified.<br /><code>@ -602,6 +633,7 </code>@ input_from_file() {<br /> get_class<br /> get_shell<br /> get_homedir<br />+ get_homeperm<br /> get_password<br /> get_expire_dates<br /> ugroups="$defaultgroups" <br /><code>@ -671,6 +703,7 </code>@ input_interactive() {<br /> get_class<br /> get_shell<br /> get_homedir<br />+ get_homeperm</p>
<pre><code>while : ; do<br /> echo <del>n "Use password-based authentication? [$_usepass]: " <br /><code>@ -781,6 +814,7 </code>@ input_interactive() {<br /> printf "%-10s : <span>s\n" "Class" "$uclass" <br /> printf "</span>-10s : <span>s %s\n" "Groups" "${ulogingroup:</del>$username}" "$ugroups" <br /> printf "</span>-10s : <span>s\n" "Home" "$uhome" <br />+ printf "</span>-10s : <span>s\n" "Home Mode" "$uhomeperm" <br /> printf "</span>-10s : <span>s\n" "Shell" "$ushell" <br /> printf "</span>-10s : %s\n" "Locked" "$_disable" <br /> while : ; do<br /><code>@ -825,6 +859,7 </code>@ ugecos=<br /> ulogingroup=<br /> uclass=<br /> uhome=<br />+uhomeperm=<br /> upass=<br /> ushell=<br /> udotdir=/usr/share/skel<br /><code>@ -851,6 +886,7 </code>@ defaultclass=<br /> defaultLgroup=<br /> defaultgroups=<br /> defaultshell="${DEFAULTSHELL}" <br />+defaultHomePerm=</code></pre>
<ol>
<li>Make sure the user running this program is root. This isn't a security</li>
<li>measure as much as it is a useful method of reminding the user to<br /><code>@ -939,6 +975,10 </code>@ for _switch ; do<br /> esac<br /> shift; shift<br /> ;;<br />+ <del>M)<br />+ defaultHomePerm=$2<br />+ shift; shift<br />+ ;;<br /> -N)<br /> readconfig=<br /> shift<br />-</del> <br />1.7.7</li>
</ol> DragonFlyBSD - Submit #2159 (Closed): [PATCH 1/2] Fix a wrong string in rmuser.shhttps://bugs.dragonflybsd.org/issues/21592011-10-26T20:11:52Zjuanfra684
<p>---<br /> usr.sbin/adduser/rmuser.sh | 2 <ins><del><br /> 1 files changed, 1 insertions(</ins>), 1 deletions(</del>)</p>
<p>diff --git a/usr.sbin/adduser/rmuser.sh b/usr.sbin/adduser/rmuser.sh<br />index fa2f4cf..91eec10 100644<br />--- a/usr.sbin/adduser/rmuser.sh<br />+<ins>+ b/usr.sbin/adduser/rmuser.sh<br /><code>@ -302,7 +302,7 </code>@ if [ ! "$userlist" ]; then<br /> show_usage<br /> exit 1<br /> else<br />- echo <del>n "Please enter one or more user name's: " <br /></ins> echo -n "Please enter one or more usernames: " <br /> read userlist<br /> fi<br /> fi<br />-</del> <br />1.7.7</p> DragonFlyBSD - Bug #2155 (Closed): libmicro doesn't work after of the big vm patchhttps://bugs.dragonflybsd.org/issues/21552011-10-22T06:46:19Zjuanfra684
<p>I was testing with libmicro the difference of performance of dragonfly 2.12<br />pre-patch vs post-patch. libmicro is a project of opensolaris for testing their<br />system against linux.</p>
<p>I ran the tests in dragonfly pre-patch and all the tests passed without<br />problems. I updated /usr/src, "make nativekernel", "make installkernel",<br />rebooted the computer and I ran the tests again.</p>
<p>Something kills the tests. I don't know what kill this. The tests die in<br />something relative to "fork".</p>
<p>My bug report is not for you fix the problem in libmicro. I think the kernel has<br />a bug relative to the vm patch and libmicro shows this.</p>
<p>The version of libmicro in pkgsrc is broken for all OS. I've reported the bug<br />(#45500). Meantime, I've made a shell script with the patches for automating the<br />process ( <a class="external" href="http://juanfra.info/bugs-y-listas/libmicro-dragonfly/libmicro.sh">http://juanfra.info/bugs-y-listas/libmicro-dragonfly/libmicro.sh</a> )<br />. Also you can download the script, to read the steps and to apply the patches<br />by hand.</p>
<p>I'm using x86 with world updated.</p> DragonFlyBSD - Bug #2153 (New): Too many unuseful warnings at boothttps://bugs.dragonflybsd.org/issues/21532011-10-19T05:16:05Zjuanfra684
<p>I have a lot of warnings at boot about the partition ad0s2 (freebsd 9 beta 3 with <br />ufs and softdep without journal).</p>
<p>If the warning is correct, show me only one warning and hide all others.</p>
<p><a class="external" href="http://juanfra.info/bugs-y-listas/pruebas-dflybsd-201110/boot1.png">http://juanfra.info/bugs-y-listas/pruebas-dflybsd-201110/boot1.png</a></p>
<p>DragonFly v2.13.0.45.g8f73b-DEVELOPMENT (GENERIC) #0: Sun Oct 16 06:54:55 UTC 2011</p> DragonFlyBSD - Bug #2151 (Resolved): DragonFly doesn't boot from usb stickhttps://bugs.dragonflybsd.org/issues/21512011-10-17T05:51:10Zjuanfra684
<p>Steps for reproduce:<br />- Download the last usb image from the mirrors.<br />- dd to your usb stick.<br />- Power off your computer.<br />- Conect two usb sticks to your computer. The first with the installer of <br />DragonFly. The seconds will be the destination of your installation.<br />- Boot the installer.<br />- Install DragonFly in your second usb stick. Use all the disk and hammer.<br />- Power off your computer.<br />- Disconnect the first usb stick.<br />- Power on your computer.<br />- Boot your usb stick.<br />- DragonFly doesn't find your root partition.</p>
<p>The old (usb) disk "da9x" is "da8x" now. Here exists two problems:<br />- In /etc/fstab the partitions point to "da9x".<br />- The same for /boot/loader.conf.</p>
<p>The installation in usb disks is useless. With this problem the user can't use <br />his or her usb disk in other computers with other order of usb disks. Also this <br />is very bad for IT departments.</p>
<p>I think the better solution is the support for UUID.</p>
<p>The same problem for other user:<br />- <a class="external" href="http://leaf.dragonflybsd.org/mailarchive/users/2011-09/msg00024.html">http://leaf.dragonflybsd.org/mailarchive/users/2011-09/msg00024.html</a></p> DragonFlyBSD - Bug #2149 (Closed): "ath_reset" messages in dmesghttps://bugs.dragonflybsd.org/issues/21492011-10-15T05:58:12Zjuanfra684
<p>Can anyone hide the constants "ath_reset" messages from kernel?.</p>
<p>The messages aren't useful for the final user, only are annoying. If you need<br />this for debug the code, you can replace this behavior by a entry in loader.conf<br />like debug.ath=1 or a especial debug mode in ifconfig or a sysctl entry or whatever.</p>
<p>Thanks :)</p>