- Devel - libvirt List Archives

[Libvir] [PATCH] lots of warnings removed - libvirt can now compile with -Werror
by Richard W.M. Jones 07 Mar '07

07 Mar '07

For a completely clean build you need to also apply the previous patch[1]. Rich. [1] https://www.redhat.com/archives/libvir-list/2007-March/msg00030.html and see also Mark's reply. -- Emerging Technologies, Red Hat http://et.redhat.com/~rjones/ 64 Baker Street, London, W1U 7DF Mobile: +44 7866 314 421 "[Negative numbers] darken the very whole doctrines of the equations and make dark of the things which are in their nature excessively obvious and simple" (Francis Maseres FRS, mathematician, 1759)

3 3

[Libvir] PATCH: Process QEMU monitor at startup
by Daniel P. Berrange 07 Mar '07

07 Mar '07

With the current QEMU driver we simply spawn the QEMU process and return to the caller immediately. If the QEMU process dies immediately for some reason the user will never no - their domain will simply never appear. To address this, the attached patch will spawn the QEMU process and then immeditely process stderr to look for the monitor PTY - or EOF indicating failure. If it finds the monitor it will also try to connect to it - if this fails in any way then the domain is terminated with extreme prejudice - we absolutely need the monitor connection for other operations to work. This should at least cause a 'Domain startup failed' message to be returned to the user although more work is needed to try and extract a useful error message from the stuff written to stderr. Also in this patch I have reversed the order in which file descriptors are added to the poll array. Currently they are added server, then client, then VMs. The trouble with this is that when processing the server event, if a new client is added to the daemon then processing of subsequent FDs is out of sync. Likewise if a client message causes a new VM to startup then processing of subsquent VM FDs is out of sync. Reversing the order trivially solves it, because VM FDs s are now processed before client FDs, and thus any new VMs started by a client don't mess things up. Likewise client FDs are processed before server FDs. A more robust long term fix would be to carry around some extra metadata instead of relying on ordering when dispatching events. Finally this patch also fixes the 'resume' operation, and allows for the VM state to be reported correctly - ie we can now report 'paused' as well as shutdown/running. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|

2 9

[Libvir] [PATCH][RFC] Add --keymap option to virt-install
by Takahashi Tomohiro 07 Mar '07

07 Mar '07

Hi, I would like to add --keymap option to virt-install in order to set up jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch (notes:It relates python-virtinst) Usage: virt-install --keymap=ja I confirmed that I can use jp106 keyboard on HVM domain. But didn't confirm PV domain. Please give me an advice, if you have it. Signed-off-by: Tomohiro Takahashi <takatom(a)jp.fujitsu.com> Thanks, Tomohiro Takahashi. ================================================================================== diff -uNrp ./libvirt.orig/src/xml.c ./libvirt/src/xml.c --- ./libvirt.orig/src/xml.c 2007-02-15 01:22:02.000000000 +0900 +++ ./libvirt/src/xml.c 2007-03-06 20:17:38.000000000 +0900 @@ -246,6 +246,7 @@ static int virDomainParseXMLGraphicsDesc xmlChar *vncport = xmlGetProp(node, BAD_CAST "port"); xmlChar *vnclisten = xmlGetProp(node, BAD_CAST "listen"); xmlChar *vncpasswd = xmlGetProp(node, BAD_CAST "passwd"); + xmlChar *keymap = xmlGetProp(node, BAD_CAST "keymap"); if (vncport != NULL) { long port = strtol((const char *)vncport, NULL, 10); if (port == -1) @@ -262,6 +263,10 @@ static int virDomainParseXMLGraphicsDesc virBufferVSprintf(buf, "(vncpasswd %s)", vncpasswd); xmlFree(vncpasswd); } + if (keymap != NULL) { + virBufferVSprintf(buf, "(keymap %s)", keymap); + xmlFree(keymap); + } } } xmlFree(graphics_type); @@ -305,6 +310,7 @@ static int virDomainParseXMLGraphicsDesc xmlChar *vncport = xmlGetProp(node, BAD_CAST "port"); xmlChar *vnclisten = xmlGetProp(node, BAD_CAST "listen"); xmlChar *vncpasswd = xmlGetProp(node, BAD_CAST "passwd"); + xmlChar *keymap = xmlGetProp(node, BAD_CAST "keymap"); if (vncport != NULL) { long port = strtol((const char *)vncport, NULL, 10); if (port == -1) @@ -321,6 +327,10 @@ static int virDomainParseXMLGraphicsDesc virBufferVSprintf(buf, "(vncpasswd %s)", vncpasswd); xmlFree(vncpasswd); } + if (keymap != NULL) { + virBufferVSprintf(buf, "(keymap %s)", keymap); + xmlFree(keymap); + } } virBufferAdd(buf, "))", 2); xmlFree(graphics_type); diff -uNrp ./libvirt.orig/src/xm_internal.c ./libvirt/src/xm_internal.c --- ./libvirt.orig/src/xm_internal.c 2007-02-23 17:51:30.000000000 +0900 +++ ./libvirt/src/xm_internal.c 2007-03-06 20:20:32.000000000 +0900 @@ -1863,6 +1863,9 @@ virConfPtr xenXMParseXMLToConfig(virConn if (xenXMConfigSetStringFromXPath(conn, conf, ctxt, "vncpasswd", "string(/domain/devices/graphics[@type='vnc']/@passwd)", 1, "cannot set the vncpasswd parameter") < 0) goto error; + if (xenXMConfigSetStringFromXPath(conn, conf, ctxt, "keymap", "string(/domain/devices/graphics[@type='vnc']/@keymap)", 1, + "cannot set the vncpasswd parameter") < 0) + goto error; /* XXX vncdisplay */ /* @@ -1894,6 +1897,7 @@ virConfPtr xenXMParseXMLToConfig(virConn xmlChar *vncport = xmlGetProp(obj->nodesetval->nodeTab[i], BAD_CAST "port"); xmlChar *vnclisten = xmlGetProp(obj->nodesetval->nodeTab[i], BAD_CAST "listen"); xmlChar *vncpasswd = xmlGetProp(obj->nodesetval->nodeTab[i], BAD_CAST "passwd"); + xmlChar *keymap = xmlGetProp(obj->nodesetval->nodeTab[i], BAD_CAST "keymap"); int vncunused = vncport ? (!strcmp((const char*)vncport, "-1") ? 1 : 0) : 1; if (vncunused) len += 12; @@ -1903,6 +1907,8 @@ virConfPtr xenXMParseXMLToConfig(virConn len += 11 + strlen((const char*)vnclisten); if (vncpasswd) len += 11 + strlen((const char*)vncpasswd); + if (keymap) + len += 11 + strlen((const char*)keymap); if ((val = malloc(len)) != NULL) { strcpy(val, "type=vnc"); if (vncunused) { @@ -1923,6 +1929,11 @@ virConfPtr xenXMParseXMLToConfig(virConn strcat(val, (const char*)vncpasswd); xmlFree(vncpasswd); } + if (keymap) { + strcat(val, ",keymap="); + strcat(val, (const char*)keymap); + xmlFree(keymap); + } } } xmlFree(type);

2 2

[Libvir] [PATCH] add global --readonly flag to virsh
by Richard W.M. Jones 07 Mar '07

07 Mar '07

The current vshInit function in virsh tries some dodgy heuristics to see if it should connect readonly or read/write to the hypervisor. Unfortunately these heuristics fail, eg. when you have a root-owned system-wide qemud, and a user trying to run virsh as non-root. This patch removes the heuristics and replaces them with a simple -r | --readonly flag on the command line. If omitted, we try to connect read/write, otherwise we try to connect readonly. Note that this doesn't affect the "connect" command in the shell, which still has its own --readonly flag that is completely separate from this new global flag. Rich.

2 2

[Libvir] [PATCH] unlink roSockname, not sockname
by Richard W.M. Jones 07 Mar '07

07 Mar '07

Very trivial fix, but necessary because it prevents the qemu proxy from restarting. Rich. -- Emerging Technologies, Red Hat http://et.redhat.com/~rjones/ 64 Baker Street, London, W1U 7DF Mobile: +44 7866 314 421 "[Negative numbers] darken the very whole doctrines of the equations and make dark of the things which are in their nature excessively obvious and simple" (Francis Maseres FRS, mathematician, 1759)

2 1

[Libvir] Error when paravirtualization domain starts with virsh.
by Michel Gauthier 07 Mar '07

07 Mar '07

With the RHEL5 RC distribution on our IA64 platform (4 CPUs), when we use virsh to start a paravirtuailzation domain, we've got the following: virsh # list Id Name State ---------------------------------- 0 Domain-0 running virsh # create hndom2 Failed to get devices for domain hndom2 error: Failed to create domain from hndom2 virsh # list Id Name State ---------------------------------- 0 Domain-0 running 24 Domain-24 paused In spite of an error message, a domain named Domain-Id is created instead of the hndom2 domain. The corresponding XML configuration file is the following: <domain type="xen"> <name>hndom2</name> <os> <type>linux</type> <kernel>/boot/efi/efi/redhat/vmlinuz-2.6.18-8.el5xen</kernel> <root>/dev/hda1</root> <initrd>/boot/efi/efi/redhat/initrd-2.6.18-8.el5xenU.img</initrd> <cmdline>console=tty0 nomca 3 ide0=noprobe ide1=noprobe ide2=noprobe ide3=noprobe selinux=0</cmdline> </os> <memory>527360</memory> <currentMemory>393216</currentMemory> <vcpu>3</vcpu> <on_poweroff>restart</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <disk type="block"> <driver name="phy"/> <source dev="/dev/sda5"/> <target dev="/dev/hda1"/> </disk> <interface type="bridge"> <source bridge=""/> </interface> </devices> </domain> If the <currentMemory> tag is suppressed from the XML configuration file, all is OK. This behaviour appears with the libvirt version 0.1.8 released with the RHEL5 RC, and also with the libvirt version 0.1.10. Are you aware of this problem? If you want some complementary infos (xend log), we can send it to you. Regards, Michel Gauthier.

3 3

[Libvir] [RFC] The idea of 'policy manager'
by ASANO Yuzuru 06 Mar '07

06 Mar '07

Hello, I summarized the idea about 'policy manager'. The purpose of 'policy manager' is to manage and to operate 'runtime policy' such as cpu pin/weight/cap. Please refer to the following thread about earlier argument about 'policy manager'. https://www.redhat.com/archives/libvir-list/2007-February/msg00211.html To implement the above mentioned features, I think that 'policy manager' need to be constructed by CLI and API. Is it all right in the following ideas? - About CLI At first, I think the following commands are necessary. ('xxx' means temporary command name) (1) xxx vcpupin <domain> <vcpu> <cpulist> : Same as virsh vcpupin <domain> <vcpu> <cpulist> command (2) xxx schedweight <domain> <weight> : Same as xm sched-credit -d <domain> -w <weight> command (3) xxx schedcap <domain> <cap> : Same as xm sched-credit -d <domain> -c <cap> command (4) xxx cpuoccupy <pcpu> <domains> : The domains specified by <domains> can use the <pcpu> in occupancy, and the other domains can't use it. (*) (5) xxx cpupolicy <domain> : Show the CPU runtime policy of <domain>. (*) When a running domain has vcpu pinned to only single pcpu, the other domains can't occupy the pcpu. In above mentioned (1)-(4), policy informations of not only running domains but also inactive domains are managed in the following API. - About API The main function which is necessary for API are as follows. o When CLI command is executed, the demanded processing is done by calling libvirt from API. o API preserves the specified by CLI command to a specific file. o When start/reboot domain command is executed by libvirt, API is called from libvirt and the domain is set the runtime policy. - TODO list o It is necessary to add new function into libvirt to change scheduler parameters (cpu weight/cap) previously. Any other thoughts from anyone? Regards, Yuzuru Asano.

1 0

[Libvir] [PATCH] qemuOpenConnection readonly parameter is not unused
by Richard W.M. Jones 06 Mar '07

06 Mar '07

2 1

[Libvir] [PATCH] patch for virsh domname <uuid> command error
by Tatsuro Enokura 02 Mar '07

02 Mar '07

Hi, Virsh domname <uuid> command for shut-off domain fails on xen-3.0.4. I think that xm_internal.c shoud be modified. But, it seems that xm_internal's structure isn't for xen-3.0.4, and shut-off domain's information is got from xend. I make a patch for xend_internal.c The attached patch resolve this issue in the following way: 1) If the domain's name can't be pulled out from the running domains list by uuid, request xend by "/xend/domains/<uuid>". 2) If the domain's data can be teken from xend, pull out the domain's name. Signed-off-by: Tatsuro Enokura <fj7716hz(a)aa.jp.fujitsu.com> Thanks, Tatsuro Enokura. Index: src/xend_internal.c (libvirt-0.2.0) =================================================================== --- xend_internal.c 2007-02-15 01:16:16.000000000 +0900 +++ xend_internal.c.new 2007-02-23 21:13:22.000000000 +0900 @@ -2759,6 +2759,8 @@ xenDaemonLookupByUUID(virConnectPtr conn char **tmp; unsigned char ident[VIR_UUID_BUFLEN]; int id = -1; + struct sexpr *root = NULL; + char *domname = NULL; names = xenDaemonListDomainsOld(conn); tmp = names; @@ -2779,6 +2781,26 @@ xenDaemonLookupByUUID(virConnectPtr conn } free(names); + if (name == NULL) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + memset(uuidstr, '\0', VIR_UUID_STRING_BUFLEN); + + snprintf(uuidstr, VIR_UUID_STRING_BUFLEN, + "%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x", + uuid[0], uuid[1], uuid[2], uuid[3], + uuid[4], uuid[5], uuid[6], uuid[7], + uuid[8], uuid[9], uuid[10], uuid[11], + uuid[12], uuid[13], uuid[14], uuid[15]); + root = sexpr_get(conn, "/xend/domain/%s?detail=1", uuidstr); + if (root == NULL) + goto error; + domname = (char*)sexpr_node(root, "domain/name"); + if (domname == NULL) + goto error; + name = strdup(domname); + id = -1; + } + if (name == NULL) goto error; @@ -2793,6 +2815,8 @@ xenDaemonLookupByUUID(virConnectPtr conn return (ret); error: + if (root != NULL) + sexpr_free(root); if (name != NULL) free(name); return (NULL); ===================================================================

2 4

[Libvir] [PATCH]Report the error from xend for Domain-0
by Kazuki　Mizushima 02 Mar '07

02 Mar '07

Hi, When I test virsh subcommands, dump save, for Domain-0, It says following normal messages, but the processing is not done. #virsh dump 0 a Domain 0 dumpd to a #echo $? 0 #virsh save 0 a Domain 0 saved to a #echo $? 0 I made a patch which reports the error from xend like this. #virsh dump 0 a libvir: Xen Daemon error : POST operation failed: (xend.err 'Cannot save privileged domain 0') libvir: error : library call virDomainSave failed, possibly not supported error: Failed to save domain 0 to a # echo $? 1 #virsh dump 0 a libvir: Xen Daemon error : POST operation failed: (xend.err 'Cannot dump core for privileged domain 0') libvir: error : library call virDomainCoreDump failed, possibly not supported error: Failed to core dump domain 0 to a # echo $? 1 Signed-off-by: Kazuki Mizushima <mizushima.kazuk(a)jp.fujitsu.com> Thanks, Kazuki Mizushima Index: xend_internal.c (libvirt-0.2.0) ---------------------------------------------------------------------- --- ../../libvirt-0.2.0/src/xend_internal.c 2007-02-15 01:16:16.000000000 +0900 +++ xend_internal.c 2007-02-27 22:45:05.000000000 +0900 @@ -507,7 +507,10 @@ xend_post(virConnectPtr xend, const char if ((ret < 0) || (ret >= 300)) { virXendError(xend, VIR_ERR_POST_FAILED, content); - } else if ((ret = 202) && (strstr(content, "failed") != NULL)) { + } else if ((ret == 202) && (strstr(content, "failed") != NULL)) { + virXendError(xend, VIR_ERR_POST_FAILED, content); + ret = -1; + } else if ((ret == 202) && (strcasestr(content, "cannot") != NULL)) { virXendError(xend, VIR_ERR_POST_FAILED, content); ret = -1; } ----------------------------------------------------------------------

4 5