Anaconda-devel December 2008

anaconda-devel@lists.fedoraproject.org

25 participants
100 discussions

[PATCH] Use macros for D-Bus paths
by David Cantrell 09 Dec '08

09 Dec '08

--- loader/net.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/loader/net.c b/loader/net.c index 69f5779..aead300 100644 --- a/loader/net.c +++ b/loader/net.c @@ -1897,7 +1897,7 @@ int get_connection(iface_t *iface) { DBusError error; DBusMessageIter iter, variant_iter; dbus_uint32_t state = 0; - char *nm_iface = "org.freedesktop.NetworkManager"; + char *nm_iface = NM_DBUS_INTERFACE; char *property = "State"; if (iface == NULL) { @@ -1931,7 +1931,7 @@ int get_connection(iface_t *iface) { dbus_error_init(&error); message = dbus_message_new_method_call(NM_DBUS_SERVICE, NM_DBUS_PATH, - "org.freedesktop.DBus.Properties", + DBUS_INTERFACE_PROPERTIES, "Get"); if (!message) { if (dbus_error_is_set(&error)) { -- 1.6.0.3

2 1

proposal - auto updates mechanism
by Chris Lumens 09 Dec '08

09 Dec '08

Instead of talking about this right before every single release, let's talk about it when there's still time to do something with it? For a while now, we've been kicking around the idea of some sort of auto updates mechanism - a way to tell anaconda to just go and get an approved updates.img with important fixes from the release. Before I get started explaining this, I'd like everyone to keep in mind that this should be completely separate from all the mirrors, ISO images, etc. This is a whole other system in order to keep bureaucracy and complexity to a minimum. I've always seen this as nothing more complicated than a directory structure on a well-known website. anaconda assembles a URL, tries to fetch the updates.img from there, and continues through to stage2 regardless of if it works or not. Let me mention a couple issues I see with supporting this, along with how I'd plan on handling them: (1) Updates can support both arch-independent Python files, as well as shared libraries. We have occassionally made updates with libraries in them. Additionally, we support several releases at a time. Our directory structure should therefore support all these possibilities. I see anaconda trying to get a very specific updates.img first and if that fails, try to get a progressively less specific image until it either finds one or runs out of options. So first it would try to get: http://wherever/f10/i386/updates.img Failing that, it tries: http://wherever/f10/updates.img, http://wherever/updates.img Once that fails, anaconda keeps going without prompting. As long as we use values we can determine from the .buildstamp or from the running system, I think this plan works out fine. (2) Getting mysterious files from a URL without asking never goes over well. I'd prefer using updates=auto as the way to kick this off since then we're not doing anything without asking. To make it easier, we could put a new entry on the boot.iso menu that includes this option. (3) We need a process for deciding which bugs are important enough for inclusion, then making sure they get some testing before being published. I'm currently adding NeedsUpdatesImg to the whiteboard and putting the bug in ON_QA when I commit a fix for it. Is this sufficient? I'd like to keep the number of fixes relatively low or this gets very complicated. Should the fixes be committed to the release's branch? How do we give them enough testing? If the fix is not obvious, does that mean it's too complicated for an update and should only be dealt with in the next major release? (4) We'll need space to store these updates images that's guaranteed to not go away. I don't really know what to say about this one. Infrastructure stuff has always been mysterious to me. What am I missing here? It doesn't sound like very much work at all on the anaconda side of things. - Chris

7 16

[PATCH] 474375 - updating instnum text
by Chris Lumens 09 Dec '08

09 Dec '08

Ignore for a minute the horrors of updating translations after a release has been made and consider the following patch. Note that running make update-po on the translation files produces a huge amount of change and introduces all the new encryption-related strings as well. We've got translators on board for doing this work, though. - Chris commit 52a3be453f98e15e442d8ad07b2c5d3ee9d0bfc9 Author: Chris Lumens <clumens(a)redhat.com> Date: Mon Dec 8 13:19:17 2008 -0500 Make the installation key text more descriptive (#474375). diff --git a/installclasses/rhel.py b/installclasses/rhel.py index 1d93189..a3fa202 100644 --- a/installclasses/rhel.py +++ b/installclasses/rhel.py @@ -48,19 +48,14 @@ class InstallClass(BaseInstallClass): } instkeyname = N_("Installation Number") - instkeydesc = N_("To install the full set of supported packages included " - "in your subscription, please enter your Installation " - "Number") - skipkeytext = N_("If you're unable to locate the Installation Number, " - "consult http://www.redhat.com/apps/support/in.html.\n\n" - "If you skip:\n" - "* You may not get access to the full set of " - "packages included in your subscription.\n" - "* It may result in an unsupported/uncertified " - "installation of Red Hat Enterprise Linux.\n" - "* You will not get software and security updates " - "for packages included in your subscription.") - + instkeydesc = N_("Would you like to enter an Installation Number " + "(sometimes called Subscription Number) now? This feature " + "enables the installer to access any extra components " + "included with your subscription. If you skip this step, " + "additional components can be installed manually later.\n\n" + "See http://www.redhat.com/InstNum/ for more information.") + skipkeytext = N_("If you cannot locate the Installation Number, consult " + "http://www.redhat.com/InstNum/") def setInstallData(self, anaconda): BaseInstallClass.setInstallData(self, anaconda)

2 1

[PATCH] 475083 - fix ejecting discs under kickstart
by Chris Lumens 08 Dec '08

08 Dec '08

Here's a stupid one. I'd like it if we had a better "/dev/whatever" vs. "whatever" policy throughout anaconda, but until that day arrives, here's another patch to handle both. - Chris diff --git a/isys/isys.py b/isys/isys.py index c7cf25b..5894617 100755 --- a/isys/isys.py +++ b/isys/isys.py @@ -788,6 +788,9 @@ def ext2HasJournal(device): return hasjournal def ejectCdrom(device): + if not os.path.exists(device): + device = "/dev/%s" % device + fd = os.open(device, os.O_RDONLY|os.O_NONBLOCK) # this is a best effort

2 1

[PATCH] 473260 - unmount swap files on upgrades
by Chris Lumens 08 Dec '08

08 Dec '08

The above referenced bug seems to be a really big deal on F10 upgrades. The problem is that swap files never get unmounted, so the filesystems they reside on never get unmounted, so migrateFilesystems fails. The attached patch corrects this by adding support to just unmount swap files. If we come to some sort of agreement on this patch, I'm in favor of including it in an updates.img. - Chris diff --git a/fsset.py b/fsset.py index a4b3cd4..d425a48 100644 --- a/fsset.py +++ b/fsset.py @@ -839,8 +839,19 @@ class swapFileSystem(FileSystemType): isys.swapon (device) def umount(self, device, path): - # unfortunately, turning off swap is bad. - raise RuntimeError, "unable to turn off swap" + # Turning off swap devices is bad, but we can get away with turning + # off swap files. + if device.device.startswith("/dev/"): + raise RuntimeError, "unable to turn off swap" + + # path is something stupid like instPath+"swap", which is totally + # useless. But we don't have instPath anywhere else so trim off + # the end of path. + swapFile = path[:path.rfind("/")] + device.device + + iutil.execWithRedirect("swapoff", [swapFile], + stdout="/dev/tty5", stderr="/dev/tty5", + searchPath=1) def formatDevice(self, entry, progress, chroot='/'): file = entry.device.setupDevice(chroot) @@ -2103,7 +2114,10 @@ MAILADDR root reverse.reverse() for entry in reverse: - if entry.mountpoint == "swap" and not swapoff: + # Always unmount swap files, or we won't be able to unmount the + # device they reside on. + if entry.mountpoint == "swap" and not swapoff and \ + entry.device.device.startswith("/dev/"): continue entry.umount(instPath) entry.device.cleanupDevice(instPath)

4 6

[PATCH] Do not write SEARCH line to ifcfg-DEVICE file (#474858)
by David Cantrell 07 Dec '08

07 Dec '08

We have been writing the SEARCH= line to ifcfg files in network.py for a while, but it only contains the domain we extract from the FQDN. For users wanting to use the network service over NM, this presents a problem because the SEARCH line present in the ifcfg file will always override the information from the DHCP server. --- network.py | 4 ---- 1 files changed, 0 insertions(+), 4 deletions(-) diff --git a/network.py b/network.py index 7becba3..b5ecad4 100644 --- a/network.py +++ b/network.py @@ -582,10 +582,6 @@ class Network: if dev.get('MTU') and dev.get('MTU') != 0: f.write("MTU=%s\n" % dev.get('MTU')) - if self.domains != ['localdomain'] and self.domains: - searchLine = string.joinfields(self.domains, ' ') - f.write("SEARCH=\"%s\"\n" % (searchLine,)) - # tell NetworkManager not to touch any interfaces used during # installation when / is on a network device. Ideally we would only # tell NM not to touch the interface(s) actually used for /, but we -- 1.6.0.3

2 3

[PATCH] In iface_ip2str(), make sure to advance to next item before continue.
by David Cantrell 07 Dec '08

07 Dec '08

We have to manually advance the iter when working with D-Bus, so if continue in a loop, be sure to call the next() function on the iter. --- isys/iface.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/isys/iface.c b/isys/iface.c index 9bc5582..d0613e2 100644 --- a/isys/iface.c +++ b/isys/iface.c @@ -224,6 +224,7 @@ char *iface_ip2str(char *ifname, int family) { -1, &error); dbus_message_unref(message); if (!devreply) { + dbus_message_iter_next(&a_iter); continue; } -- 1.6.0.3

2 2

[PATCH] 470253 - remove markup from cmdline mode
by Chris Lumens 07 Dec '08

07 Dec '08

This patch removes the <b></b> markup from package installation in cmdline mode, just like we do for text installs. - Chris

3 5

[PATCH] Use strverscmp() from glibc in place of rpmvercmp()
by David Cantrell 07 Dec '08

07 Dec '08

We are linking with glibc now, so use strverscmp() in place of the rpmvercmp() function in loadermisc.c. --- loader/Makefile | 2 +- loader/dirbrowser.c | 2 +- loader/loadermisc.c | 110 +++----------------------------------------------- 3 files changed, 9 insertions(+), 105 deletions(-) diff --git a/loader/Makefile b/loader/Makefile index e83d0f2..25dfb64 100644 --- a/loader/Makefile +++ b/loader/Makefile @@ -65,7 +65,7 @@ SOURCES = $(subst .o,.c,$(OBJS)) loader.c LIBS += -CFLAGS += -DUSE_LOGDEV -DVERSION='"$(VERSION)"' +CFLAGS += -DUSE_LOGDEV -DVERSION='"$(VERSION)"' -D_GNU_SOURCE REALCC=gcc # linuxrc + shutdown on s390, init everywhere else diff --git a/loader/dirbrowser.c b/loader/dirbrowser.c index 4d9c25c..6c713ec 100644 --- a/loader/dirbrowser.c +++ b/loader/dirbrowser.c @@ -43,7 +43,7 @@ static int simpleStringCmp(const void * a, const void * b) { const char * first = *((const char **) a); const char * second = *((const char **) b); - + return strcmp(first, second); } #endif diff --git a/loader/loadermisc.c b/loader/loadermisc.c index a156007..03d26bb 100644 --- a/loader/loadermisc.c +++ b/loader/loadermisc.c @@ -86,107 +86,11 @@ int copyFile(char * source, char * dest) { return rc; } -/* FIXME: when we only depend on glibc, we could use strvercmp instead */ -/* compare alpha and numeric segments of two versions */ -/* return 1: a is newer than b */ -/* 0: a and b are the same version */ -/* -1: b is newer than a */ -static int rpmvercmp(const char * a, const char * b) -{ - char oldch1, oldch2; - char * str1, * str2; - char * one, * two; - int rc; - int isnum; - - /* easy comparison to see if versions are identical */ - if (!strcmp(a, b)) return 0; - - str1 = alloca(strlen(a) + 1); - str2 = alloca(strlen(b) + 1); - - strcpy(str1, a); - strcpy(str2, b); - - one = str1; - two = str2; - - /* loop through each version segment of str1 and str2 and compare them */ - while (*one && *two) { - while (*one && !isalnum(*one)) one++; - while (*two && !isalnum(*two)) two++; - - str1 = one; - str2 = two; - - /* grab first completely alpha or completely numeric segment */ - /* leave one and two pointing to the start of the alpha or numeric */ - /* segment and walk str1 and str2 to end of segment */ - if (isdigit(*str1)) { - while (*str1 && isdigit(*str1)) str1++; - while (*str2 && isdigit(*str2)) str2++; - isnum = 1; - } else { - while (*str1 && isalpha(*str1)) str1++; - while (*str2 && isalpha(*str2)) str2++; - isnum = 0; - } - - /* save character at the end of the alpha or numeric segment */ - /* so that they can be restored after the comparison */ - oldch1 = *str1; - *str1 = '\0'; - oldch2 = *str2; - *str2 = '\0'; - - /* take care of the case where the two version segments are */ - /* different types: one numeric, the other alpha (i.e. empty) */ - if (one == str1) return -1; /* arbitrary */ - /* XXX See patch #60884 (and details) from bugzilla #50977. */ - if (two == str2) return (isnum ? 1 : -1); - - if (isnum) { - /* this used to be done by converting the digit segments */ - /* to ints using atoi() - it's changed because long */ - /* digit segments can overflow an int - this should fix that. */ - - /* throw away any leading zeros - it's a number, right? */ - while (*one == '0') one++; - while (*two == '0') two++; - - /* whichever number has more digits wins */ - if (strlen(one) > strlen(two)) return 1; - if (strlen(two) > strlen(one)) return -1; - } - - /* strcmp will return which one is greater - even if the two */ - /* segments are alpha or if they are numeric. don't return */ - /* if they are equal because there might be more segments to */ - /* compare */ - rc = strcmp(one, two); - if (rc) return (rc < 1 ? -1 : 1); - - /* restore character that was replaced by null above */ - *str1 = oldch1; - one = str1; - *str2 = oldch2; - two = str2; - } - - /* this catches the case where all numeric and alpha segments have */ - /* compared identically but the segment sepparating characters were */ - /* different */ - if ((!*one) && (!*two)) return 0; - - /* whichever version still has characters left over wins */ - if (!*one) return -1; else return 1; -} - int simpleStringCmp(const void * a, const void * b) { const char * first = *((const char **) a); const char * second = *((const char **) b); - return rpmvercmp(first, second); + return strverscmp(first, second); } /* look for available memory. note: won't ever report more than the @@ -197,23 +101,23 @@ int totalMemory(void) { char buf[4096]; char * chptr, * start; int total = 0; - + fd = open("/proc/meminfo", O_RDONLY); if (fd < 0) { logMessage(ERROR, "failed to open /proc/meminfo: %m"); return 0; } - + bytesRead = read(fd, buf, sizeof(buf) - 1); if (bytesRead < 0) { logMessage(ERROR, "failed to read from /proc/meminfo: %m"); close(fd); return 0; } - + close(fd); buf[bytesRead] = '\0'; - + chptr = buf; while (*chptr && !total) { if (strncmp(chptr, "MemTotal:", 9)) { @@ -225,7 +129,7 @@ int totalMemory(void) { while (*chptr && *chptr != '\n') chptr++; *chptr = '\0'; - + while (!isdigit(*start) && *start) start++; if (!*start) { logMessage(WARNING, "no number appears after MemTotal tag"); @@ -240,6 +144,6 @@ int totalMemory(void) { } logMessage(INFO, "%d kB are available", total); - + return total; } -- 1.6.0.3

2 2

[PATCH] We already have _GNU_SOURCE defined in Makefile.inc
by David Cantrell 07 Dec '08

07 Dec '08

Do not define _GNU_SOURCE in loader/Makefile again. --- loader/Makefile | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/loader/Makefile b/loader/Makefile index 25dfb64..e83d0f2 100644 --- a/loader/Makefile +++ b/loader/Makefile @@ -65,7 +65,7 @@ SOURCES = $(subst .o,.c,$(OBJS)) loader.c LIBS += -CFLAGS += -DUSE_LOGDEV -DVERSION='"$(VERSION)"' -D_GNU_SOURCE +CFLAGS += -DUSE_LOGDEV -DVERSION='"$(VERSION)"' REALCC=gcc # linuxrc + shutdown on s390, init everywhere else -- 1.6.0.3

2 1

← Newer
1
2
3
4
5
6
7
8
9
10
Older →

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel December 2008