From: David Gibson <david@gibson.dropbear.id.au>
To: Bharata B Rao <bharata@linux.vnet.ibm.com>
Cc: qemu-devel@nongnu.org, qemu-ppc@nongnu.org, nfont@linux.vnet.ibm.com
Subject: Re: [Qemu-devel] [FIX PATCH v0] spapr: Allow configure-connector to be called multiple times for LMBs
Date: Thu, 17 Aug 2017 11:29:19 +1000 [thread overview]
Message-ID: <20170817012919.GA5509@umbus.fritz.box> (raw)
In-Reply-To: <1502883336-14608-1-git-send-email-bharata@linux.vnet.ibm.com>
[-- Attachment #1: Type: text/plain, Size: 4503 bytes --]
On Wed, Aug 16, 2017 at 05:05:36PM +0530, Bharata B Rao wrote:
> In case of in-kernel memory hot unplug, when the guest is not able
> to remove all the LMBs that are requested for removal, it will add back
> any LMBs that have been successfully removed. The DR Connectors of
> these LMBs wouldn't have been unconfigured and hence the addition of
> these LMBs will result in configure-connector call being issued on
> LMB DR connectors that are already in configured state. Such
> configure-connector calls will fail resulting in a DIMM which is
> partially unplugged.
>
> This however worked till recently before we overhauled the DRC
> implementation in QEMU. Commit 9d4c0f4f0a71e: "spapr: Consolidate
> DRC state variables" is the first commit where this problem shows up
> as per git bisect.
>
> Ideally guest shouldn't be issuing configure-connector call on an
> already configured DR connector. However for now, work around this in
> QEMU by allowing configure-connector to be called multiple times for
> LMBs.
>
> Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
> ---
> hw/ppc/spapr_drc.c | 37 +++++++++++++++++++++++++++++++------
> 1 file changed, 31 insertions(+), 6 deletions(-)
>
> diff --git a/hw/ppc/spapr_drc.c b/hw/ppc/spapr_drc.c
> index 5260b5d..2dd9635 100644
> --- a/hw/ppc/spapr_drc.c
> +++ b/hw/ppc/spapr_drc.c
> @@ -446,8 +446,17 @@ void spapr_drc_reset(sPAPRDRConnector *drc)
> drc->state = drck->empty_state;
> }
>
> - drc->ccs_offset = -1;
> - drc->ccs_depth = -1;
> + if (spapr_drc_type(drc) == SPAPR_DR_CONNECTOR_TYPE_LMB) {
We shouldn't have a type dependency here. If we want the configure
connector process to be repeatable, it should be repeatable for
everything, even if we only actually need it for LMBs.
> + /*
> + * Ensure that we are able to send the FDT fragment of the
> + * LMB again via configure-connector call if guest requests.
> + */
> + drc->ccs_offset = drc->fdt_start_offset;
> + drc->ccs_depth = 0;
> + } else {
> + drc->ccs_offset = -1;
> + drc->ccs_depth = -1;
> + }
> }
>
> static void drc_reset(void *opaque)
> @@ -1071,8 +1080,14 @@ static void rtas_ibm_configure_connector(PowerPCCPU *cpu,
> }
>
> if ((drc->state != SPAPR_DRC_STATE_LOGICAL_UNISOLATE)
> - && (drc->state != SPAPR_DRC_STATE_PHYSICAL_UNISOLATE)) {
> - /* Need to unisolate the device before configuring */
> + && (drc->state != SPAPR_DRC_STATE_PHYSICAL_UNISOLATE) &&
> + (spapr_drc_type(drc) != SPAPR_DR_CONNECTOR_TYPE_LMB)) {
> + /*
> + * Need to unisolate the device before configuring, however
> + * LMB DRCs are exempted from this check as guest can issue
> + * configure-connector calls for an already configured
> + * LMB DRC.
> + */
Same here - but we do need to explicitly check that the state is
either UNISOLATE *or* CONFIGURED, and not allow it from some other
random state.
> rc = SPAPR_DR_CC_RESPONSE_NOT_CONFIGURABLE;
> goto out;
> }
> @@ -1108,8 +1123,18 @@ static void rtas_ibm_configure_connector(PowerPCCPU *cpu,
> /* done sending the device tree, move to configured state */
> trace_spapr_drc_set_configured(drc_index);
> drc->state = drck->ready_state;
> - drc->ccs_offset = -1;
> - drc->ccs_depth = -1;
> + if (spapr_drc_type(drc) == SPAPR_DR_CONNECTOR_TYPE_LMB) {
And again.
> + /*
> + * Ensure that we are able to send the FDT fragment of the
> + * LMB again via configure-connector call if guest requests.
> + */
> + drc->ccs_offset = drc->fdt_start_offset;
> + drc->ccs_depth = 0;
> + fdt_offset_next = drc->fdt_start_offset;
> + } else {
> + drc->ccs_offset = -1;
> + drc->ccs_depth = -1;
> + }
> resp = SPAPR_DR_CC_RESPONSE_SUCCESS;
> } else {
> resp = SPAPR_DR_CC_RESPONSE_PREV_PARENT;
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
prev parent reply other threads:[~2017-08-17 1:34 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-16 11:35 [Qemu-devel] [FIX PATCH v0] spapr: Allow configure-connector to be called multiple times for LMBs Bharata B Rao
2017-08-17 1:29 ` David Gibson [this message]
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=20170817012919.GA5509@umbus.fritz.box \
--to=david@gibson.dropbear.id.au \
--cc=bharata@linux.vnet.ibm.com \
--cc=nfont@linux.vnet.ibm.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).