From: Jeremy Fitzhardinge <jeremy@goop.org>
To: Samuel Thibault <samuel.thibault@ens-lyon.org>,
xen-devel@lists.xensource.com
Subject: Re: [PATCH] mini-os: Fix xenbus initialisation
Date: Thu, 08 Apr 2010 12:13:41 -0700 [thread overview]
Message-ID: <4BBE2AE5.9050504@goop.org> (raw)
In-Reply-To: <20100405232748.GK23034@const.famille.thibault.fr>
On 04/05/2010 04:27 PM, Samuel Thibault wrote:
> mini-os: Fix xenbus initialisation
>
> This fixes xenbus initialization of blkfront, netfront and pcifront
> by uniformizing with fbfront: after writing parameters, set state to
> initialised, then wait for backend to switch to connect state, and then
> only read its parameter and switch to the connect state.
>
This change causes a regression for stub domains. When this is applied,
the domain and stub come up, but the main domain just sits there - not
running, not blocked. The vnc window remains black.
sh-4.0# xm destroy debhv32 && sleep 2; xm create debhv32
Using config file "./debhv32".
Started domain debhv32 (id=19)
sh-4.0# xm list
Name ID Mem VCPUs State Time(s)
Domain-0 0 2140 4 r----- 4945.7
debhv32 19 512 1 ------ 0.0
debhv32-dm 20 32 1 -b---- 0.0
The Xen console shows:
blktap_sysfs_create: adding attributes for dev ffff8800a1a07c00
blktap_control_allocate_tap: allocated tap ffff8800bbcb8800
blktap_ring_open: opening device blktap0
blktap_ring_open: opened device 0
blktap_ring_mmap: blktap: mapping pid is 402
blktap_validate_params: aio:/dev/vg_lilith-raid/xen-debian-32: capacity: 20971520, sector-size: 512
blktap_validate_params: aio:/dev/vg_lilith-raid/xen-debian-32: capacity: 20971520, sector-size: 512
blktap_device_create: minor 0 sectors 20971520 sector-size 512
device vif19.0 entered promiscuous mode
eth0: port 2(vif19.0) entering forwarding state
device vif20.0 entered promiscuous mode
eth0: port 3(vif20.0) entering forwarding state
Bootstrapping...
Xen Minimal OS!
start_info: 0x563000(VA)
nr_pages: 0x2000
shared_inf: 0xb81b1000(MA)
pt_base: 0x566000(VA)
nr_pt_frames: 0x7
mfn_list: 0x553000(VA)
mod_start: 0x0(VA)
mod_len: 0
flags: 0x0
cmd_line: -d 19
stack: 0x511700-0x531700
MM: Init
_text: 0x0(VA)
_etext: 0xffa73(VA)
_erodata: 0x150000(VA)
_edata: 0x158cc8(VA)
stack start: 0x511700(VA)
_end: 0x552010(VA)
start_pfn: 570
max_pfn: 2000
Mapping memory range 0x800000 - 0x2000000
setting 0x0-0x150000 readonly
skipped 0x1000
MM: Initialise page allocator for 57c000(57c000)-2000000(2000000)
MM: done
Demand map pfns at 2001000-2002001000.
Heap resides at 2002002000-4002002000.
Initialising timer interface
Initialising console ... done.
gnttab_table mapped at 0x2001000.
Initialising scheduler
Thread "Idle": pointer: 0x2002002050, stack: 0x5a0000
Initialising xenbus
Thread "xenstore": pointer: 0x2002002800, stack: 0x5b0000
Dummy main: start_info=0x531800
Thread "main": pointer: 0x2002002fb0, stack: 0x5c0000
Initing FS frontend(s).
No FS import
Thread "pcifront": pointer: 0x2002003760, stack: 0x5d0000
pcifront_watches: waiting for backend path to appear device/pci/0/backend
dom vm is at /vm/f2249d1c-5862-444d-a57a-c3c8a6243d1f
************************ NETFRONT for device/vif/0 **********
net TX ring size 256
net RX ring size 256
backend at /local/domain/0/backend/vif/20/0
mac is 00:16:3e:24:3d:1f
vif19.0: no IPv6 routers present
(XEN) HVM19: HVM Loader
(XEN) HVM19: Detected Xen v4.1-unstable
(XEN) HVM19: CPU speed is 2400 MHz
vif20.0: no IPv6 routers present
Reverting this change (b20f897d6010) makes hvm+stubdom work again.
Thanks,
J
> Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
>
> diff -r ebd84be3420a extras/mini-os/blkfront.c
> --- a/extras/mini-os/blkfront.c Tue Mar 30 18:31:39 2010 +0100
> +++ b/extras/mini-os/blkfront.c Tue Apr 06 01:24:28 2010 +0200
> @@ -152,13 +152,12 @@
> }
>
> snprintf(path, sizeof(path), "%s/state", nodename);
> - err = xenbus_switch_state(xbt, path, XenbusStateConnected);
> + err = xenbus_switch_state(xbt, path, XenbusStateInitialised);
> if (err) {
> - message = "switching state";
> + printk("error writing blk initialized on %s: %s\n", path, err);
> goto abort_transaction;
> }
>
> -
> err = xenbus_transaction_end(xbt, 0, &retry);
> if (err) free(err);
> if (retry) {
> @@ -189,6 +188,7 @@
> {
> XenbusState state;
> char path[strlen(dev->backend) + 1 + 19 + 1];
> + char frontpath[strlen(nodename) + 1 + 6 + 1];
> snprintf(path, sizeof(path), "%s/mode", dev->backend);
> msg = xenbus_read(XBT_NIL, path, &c);
> if (msg) {
> @@ -232,7 +232,18 @@
> dev->info.flush = xenbus_read_integer(path);
>
> *info = dev->info;
> +
> + printk("%s connected\n", dev->backend);
> +
> + snprintf(frontpath, sizeof(frontpath), "%s/state", nodename);
> + if((err = xenbus_switch_state(XBT_NIL, frontpath, XenbusStateConnected))
> + != NULL) {
> + printk("error switching state: %s\n", err);
> + xenbus_unwatch_path_token(XBT_NIL, path, path);
> + goto error;
> + }
> }
> +
> unmask_evtchn(dev->evtchn);
>
> printk("%u sectors of %u bytes\n", dev->info.sectors, dev->info.sector_size);
> diff -r ebd84be3420a extras/mini-os/fbfront.c
> --- a/extras/mini-os/fbfront.c Tue Mar 30 18:31:39 2010 +0100
> +++ b/extras/mini-os/fbfront.c Tue Apr 06 01:24:28 2010 +0200
> @@ -126,8 +126,8 @@
> snprintf(path, sizeof(path), "%s/state", nodename);
> err = xenbus_switch_state(xbt, path, XenbusStateInitialised);
> if (err) {
> - printk("error writing initialized: %s\n", err);
> - free(err);
> + printk("error writing fb initialized: %s\n", err);
> + goto abort_transaction;
> }
>
> err = xenbus_transaction_end(xbt, 0, &retry);
> @@ -189,6 +189,7 @@
> printk("************************** KBDFRONT\n");
>
> return dev;
> +
> error:
> free(msg);
> free(err);
> diff -r ebd84be3420a extras/mini-os/netfront.c
> --- a/extras/mini-os/netfront.c Tue Mar 30 18:31:39 2010 +0100
> +++ b/extras/mini-os/netfront.c Tue Apr 06 01:24:28 2010 +0200
> @@ -407,9 +407,9 @@
> }
>
> snprintf(path, sizeof(path), "%s/state", nodename);
> - err = xenbus_switch_state(xbt, path, XenbusStateConnected);
> + err = xenbus_switch_state(xbt, path, XenbusStateInitialised);
> if (err) {
> - message = "switching state";
> + printk("error writing net initialized: %s\n", err);
> goto abort_transaction;
> }
>
> @@ -445,6 +445,7 @@
> {
> XenbusState state;
> char path[strlen(dev->backend) + 1 + 5 + 1];
> + char frontpath[strlen(nodename) + 1 + 6 + 1];
> snprintf(path, sizeof(path), "%s/state", dev->backend);
>
> xenbus_watch_path_token(XBT_NIL, path, path, &dev->events);
> @@ -463,6 +464,16 @@
> snprintf(path, sizeof(path), "%s/ip", dev->backend);
> xenbus_read(XBT_NIL, path, ip);
> }
> +
> + printk("%s connected\n", dev->backend);
> +
> + snprintf(frontpath, sizeof(frontpath), "%s/state", nodename);
> + if((err = xenbus_switch_state(XBT_NIL, frontpath, XenbusStateConnected))
> + != NULL) {
> + printk("error switching state: %s\n", err);
> + xenbus_unwatch_path_token(XBT_NIL, path, path);
> + goto error;
> + }
> }
>
> printk("**************************\n");
> @@ -481,6 +492,7 @@
> &rawmac[5]);
>
> return dev;
> +
> error:
> free(msg);
> free(err);
> diff -r ebd84be3420a extras/mini-os/pcifront.c
> --- a/extras/mini-os/pcifront.c Tue Mar 30 18:31:39 2010 +0100
> +++ b/extras/mini-os/pcifront.c Tue Apr 06 01:24:28 2010 +0200
> @@ -143,7 +143,7 @@
> char* err;
> char* message=NULL;
> int retry=0;
> - char* msg;
> + char* msg = NULL;
> char* nodename = _nodename ? _nodename : "device/pci/0";
> int dom;
>
> @@ -206,7 +206,7 @@
> snprintf(path, sizeof(path), "%s/state", nodename);
> err = xenbus_switch_state(xbt, path, XenbusStateInitialised);
> if (err) {
> - message = "switching state";
> + printk("error writing pci initialized: %s\n", err);
> goto abort_transaction;
> }
>
> @@ -237,7 +237,7 @@
>
> {
> char path[strlen(dev->backend) + 1 + 5 + 1];
> - char frontpath[strlen(nodename) + 1 + 5 + 1];
> + char frontpath[strlen(nodename) + 1 + 6 + 1];
> XenbusState state;
> snprintf(path, sizeof(path), "%s/state", dev->backend);
>
> @@ -254,9 +254,9 @@
> }
>
> snprintf(frontpath, sizeof(frontpath), "%s/state", nodename);
> - if ((err = xenbus_switch_state(XBT_NIL, frontpath, XenbusStateConnected))
> + if((err = xenbus_switch_state(XBT_NIL, frontpath, XenbusStateConnected))
> != NULL) {
> - printk("error switching state %s\n", err);
> + printk("error switching state: %s\n", err);
> xenbus_unwatch_path_token(XBT_NIL, path, path);
> goto error;
> }
> @@ -271,6 +271,7 @@
> return dev;
>
> error:
> + free(msg);
> free(err);
> free_pcifront(dev);
> return NULL;
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>
>
next prev parent reply other threads:[~2010-04-08 19:13 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-04-05 23:27 [PATCH] mini-os: Fix xenbus initialisation Samuel Thibault
2010-04-08 19:13 ` Jeremy Fitzhardinge [this message]
2010-04-08 21:25 ` Samuel Thibault
2010-04-08 22:39 ` Jeremy Fitzhardinge
2010-04-29 20:36 ` Jeremy Fitzhardinge
2010-05-12 0:33 ` Samuel Thibault
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4BBE2AE5.9050504@goop.org \
--to=jeremy@goop.org \
--cc=samuel.thibault@ens-lyon.org \
--cc=xen-devel@lists.xensource.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.