[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-ppc] [RFC PATCH v2 12/21] ppc/xive: notify the CPU when interr
From: |
David Gibson |
Subject: |
Re: [Qemu-ppc] [RFC PATCH v2 12/21] ppc/xive: notify the CPU when interrupt priority is more privileged |
Date: |
Tue, 19 Sep 2017 17:50:08 +1000 |
User-agent: |
Mutt/1.8.3 (2017-05-23) |
On Mon, Sep 11, 2017 at 07:12:26PM +0200, Cédric Le Goater wrote:
> The Pending Interrupt Priority Register (PIPR) contains the priority
> of the most favored pending notification. It is calculated from the
> Interrupt Pending Buffer (IPB) which indicates a pending interrupt at
> the priority corresponding to the bit number.
>
> If the PIPR is more favored (1) than the Current Processor Priority
> Register (CPPR), the CPU interrupt line can be raised and the EO bit
> of the Notification Source Register is updated to notify the presence
> of an exception for the O/S. The check needs to be done whenever the
> PIPR or the CPPR is changed.
>
> (1) numerically less than
>
> Signed-off-by: Cédric Le Goater <address@hidden>
> ---
> hw/intc/spapr_xive.c | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 50 insertions(+)
>
> diff --git a/hw/intc/spapr_xive.c b/hw/intc/spapr_xive.c
> index 4bc61cfda67a..e5d4b723b7e0 100644
> --- a/hw/intc/spapr_xive.c
> +++ b/hw/intc/spapr_xive.c
> @@ -28,11 +28,39 @@
> #include "xive-internal.h"
>
>
> +/* Convert a priority number to an Interrupt Pending Buffer (IPB)
> + * register, which indicates a pending interrupt at the priority
> + * corresponding to the bit number
> + */
> +static uint8_t priority_to_ipb(uint8_t priority)
> +{
> + return priority > XIVE_PRIORITY_MAX ? 0 : 1 << (7 - priority);
> +}
> +
> +/* Convert an Interrupt Pending Buffer (IPB) register to a Pending
> + * Interrupt Priority Register (PIPR), which contains the priority of
> + * the most favored pending notification.
> + *
> + * TODO: PIPR can never be OxFF. Needs a fix.
> + */
> +static uint8_t ipb_to_pipr(uint8_t ibp)
> +{
> + return ibp ? clz32((uint32_t)ibp << 24) : 0xff;
> +}
> +
> static uint64_t spapr_xive_icp_accept(ICPState *icp)
> {
> return 0;
> }
>
> +static void spapr_xive_icp_notify(ICPState *icp)
> +{
> + if (icp->tima_os[TM_PIPR] < icp->tima_os[TM_CPPR]) {
> + icp->tima_os[TM_NSR] |= TM_QW1_NSR_EO;
> + qemu_irq_raise(ICP(icp)->output);
The CPU interrupt lines are effectively level sensitive, but you never
lower this, AFAICT.
> + }
> +}
> +
> static void spapr_xive_icp_set_cppr(ICPState *icp, uint8_t cppr)
> {
> if (cppr > XIVE_PRIORITY_MAX) {
> @@ -40,6 +68,10 @@ static void spapr_xive_icp_set_cppr(ICPState *icp, uint8_t
> cppr)
> }
>
> icp->tima_os[TM_CPPR] = cppr;
> +
> + /* CPPR has changed, inform the ICP which might raise an
> + * exception */
> + spapr_xive_icp_notify(icp);
> }
>
> /*
> @@ -206,6 +238,8 @@ static void spapr_xive_irq(sPAPRXive *xive, int srcno)
> XiveEQ *eq;
> uint32_t eq_idx;
> uint32_t priority;
> + uint32_t target;
> + ICPState *icp;
>
> ive = spapr_xive_get_ive(xive, srcno);
> if (!ive || !(ive->w & IVE_VALID)) {
> @@ -235,6 +269,13 @@ static void spapr_xive_irq(sPAPRXive *xive, int srcno)
> qemu_log_mask(LOG_UNIMP, "XIVE: !UCOND_NOTIFY not implemented\n");
> }
>
> + target = GETFIELD(EQ_W6_NVT_INDEX, eq->w6);
> + icp = xics_icp_get(xive->ics->xics, target);
> + if (!icp) {
> + qemu_log_mask(LOG_GUEST_ERROR, "XIVE: No ICP for target %d\n",
> target);
> + return;
> + }
> +
> if (GETFIELD(EQ_W6_FORMAT_BIT, eq->w6) == 0) {
> priority = GETFIELD(EQ_W7_F0_PRIORITY, eq->w7);
>
> @@ -242,9 +283,18 @@ static void spapr_xive_irq(sPAPRXive *xive, int srcno)
> if (priority == 0xff) {
> return;
> }
> +
> + /* Update the IPB (Interrupt Pending Buffer) with the priority
> + * of the new notification and inform the ICP, which will
> + * decide to raise the exception, or not, depending the CPPR.
> + */
> + icp->tima_os[TM_IPB] |= priority_to_ipb(priority);
> + icp->tima_os[TM_PIPR] = ipb_to_pipr(icp->tima_os[TM_IPB]);
> } else {
> qemu_log_mask(LOG_UNIMP, "XIVE: w7 format1 not implemented\n");
> }
> +
> + spapr_xive_icp_notify(icp);
> }
>
> /*
--
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
signature.asc
Description: PGP signature
- Re: [Qemu-ppc] [RFC PATCH v2 09/21] ppc/xive: extend the interrupt presenter model for XIVE, (continued)
[Qemu-ppc] [RFC PATCH v2 10/21] ppc/xive: add MMIO handlers for the XIVE TIMA, Cédric Le Goater, 2017/09/11
[Qemu-ppc] [RFC PATCH v2 11/21] ppc/xive: push the EQ data in OS event queue, Cédric Le Goater, 2017/09/11
[Qemu-ppc] [RFC PATCH v2 12/21] ppc/xive: notify the CPU when interrupt priority is more privileged, Cédric Le Goater, 2017/09/11
- Re: [Qemu-ppc] [RFC PATCH v2 12/21] ppc/xive: notify the CPU when interrupt priority is more privileged,
David Gibson <=
[Qemu-ppc] [RFC PATCH v2 13/21] ppc/xive: handle interrupt acknowledgment by the O/S, Cédric Le Goater, 2017/09/11
[Qemu-ppc] [RFC PATCH v2 14/21] ppc/xive: add support for the SET_OS_PENDING command, Cédric Le Goater, 2017/09/11
[Qemu-ppc] [RFC PATCH v2 15/21] spapr: modify spapr_populate_pci_dt() to use a 'nr_irqs' argument, Cédric Le Goater, 2017/09/11