qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Cédric Le Goater" <clg@kaod.org>
To: David Gibson <david@gibson.dropbear.id.au>
Cc: qemu-ppc@nongnu.org, qemu-devel@nongnu.org,
	"Benjamin Herrenschmidt" <benh@kernel.crashing.org>,
	"Cédric Le Goater" <clg@kaod.org>
Subject: [Qemu-devel] [PATCH 02/10] target/ppc: Don't clobber MSR:EE on PM instructions
Date: Fri, 15 Feb 2019 17:16:40 +0100	[thread overview]
Message-ID: <20190215161648.9600-3-clg@kaod.org> (raw)
In-Reply-To: <20190215161648.9600-1-clg@kaod.org>

From: Benjamin Herrenschmidt <benh@kernel.crashing.org>

When issuing a power management instruction, we set MSR:EE
to force ppc_hw_interrupt() into calling powerpc_excp()
to deal with the fact that on P7 and P8, the system reset
caused by the wakeup needs to be generated regardless of
the MSR:EE value (using LPCR only).

This however means that the OS will see a bogus SRR1:EE
value which is a problem. It also prevents properly
implementing P9 STOP "light".

So fix this by instead putting some logic in ppc_hw_interrupt()
to decide whether to deliver or not by taking into account the
fact that we are waking up from sleep.

The LPCR isn't checked as this is done in the has_work() test.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
---
 target/ppc/excp_helper.c | 27 +++++++++++++++------------
 1 file changed, 15 insertions(+), 12 deletions(-)

diff --git a/target/ppc/excp_helper.c b/target/ppc/excp_helper.c
index 8407e0ade938..7c7c8d1b9dc6 100644
--- a/target/ppc/excp_helper.c
+++ b/target/ppc/excp_helper.c
@@ -748,6 +748,7 @@ void ppc_cpu_do_interrupt(CPUState *cs)
 static void ppc_hw_interrupt(CPUPPCState *env)
 {
     PowerPCCPU *cpu = ppc_env_get_cpu(env);
+    bool async_deliver;
 
     /* External reset */
     if (env->pending_interrupts & (1 << PPC_INTERRUPT_RESET)) {
@@ -769,11 +770,20 @@ static void ppc_hw_interrupt(CPUPPCState *env)
         return;
     }
 #endif
+
+    /*
+     * For interrupts that gate on MSR:EE, we need to do something a
+     * bit more subtle, as we need to let them through even when EE is
+     * clear when coming out of some power management states (in order
+     * for them to become a 0x100).
+     */
+    async_deliver = (msr_ee != 0) || env->in_pm_state;
+
     /* Hypervisor decrementer exception */
     if (env->pending_interrupts & (1 << PPC_INTERRUPT_HDECR)) {
         /* LPCR will be clear when not supported so this will work */
         bool hdice = !!(env->spr[SPR_LPCR] & LPCR_HDICE);
-        if ((msr_ee != 0 || msr_hv == 0) && hdice) {
+        if ((async_deliver || msr_hv == 0) && hdice) {
             /* HDEC clears on delivery */
             env->pending_interrupts &= ~(1 << PPC_INTERRUPT_HDECR);
             powerpc_excp(cpu, env->excp_model, POWERPC_EXCP_HDECR);
@@ -783,7 +793,7 @@ static void ppc_hw_interrupt(CPUPPCState *env)
     /* Extermal interrupt can ignore MSR:EE under some circumstances */
     if (env->pending_interrupts & (1 << PPC_INTERRUPT_EXT)) {
         bool lpes0 = !!(env->spr[SPR_LPCR] & LPCR_LPES0);
-        if (msr_ee != 0 || (env->has_hv_mode && msr_hv == 0 && !lpes0)) {
+        if (async_deliver || (env->has_hv_mode && msr_hv == 0 && !lpes0)) {
             powerpc_excp(cpu, env->excp_model, POWERPC_EXCP_EXTERNAL);
             return;
         }
@@ -795,7 +805,7 @@ static void ppc_hw_interrupt(CPUPPCState *env)
             return;
         }
     }
-    if (msr_ee != 0) {
+    if (async_deliver != 0) {
         /* Watchdog timer on embedded PowerPC */
         if (env->pending_interrupts & (1 << PPC_INTERRUPT_WDT)) {
             env->pending_interrupts &= ~(1 << PPC_INTERRUPT_WDT);
@@ -943,21 +953,14 @@ void helper_pminsn(CPUPPCState *env, powerpc_pm_insn_t insn)
 
     cs = CPU(ppc_env_get_cpu(env));
     cs->halted = 1;
-    env->in_pm_state = true;
 
     /* The architecture specifies that HDEC interrupts are
      * discarded in PM states
      */
     env->pending_interrupts &= ~(1 << PPC_INTERRUPT_HDECR);
 
-    /* Technically, nap doesn't set EE, but if we don't set it
-     * then ppc_hw_interrupt() won't deliver. We could add some
-     * other tests there based on LPCR but it's simpler to just
-     * whack EE in. It will be cleared by the 0x100 at wakeup
-     * anyway. It will still be observable by the guest in SRR1
-     * but this doesn't seem to be a problem.
-     */
-    env->msr |= (1ull << MSR_EE);
+    /* Condition for waking up at 0x100 */
+    env->in_pm_state = true;
 }
 #endif /* defined(TARGET_PPC64) */
 
-- 
2.20.1

  parent reply	other threads:[~2019-02-15 16:17 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-02-15 16:16 [Qemu-devel] [PATCH 00/10] ppc: Add Hypervisor Virtualization Interrupt on POWER9 Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 01/10] target/ppc: Fix nip on power management instructions Cédric Le Goater
2019-02-15 16:16 ` Cédric Le Goater [this message]
2019-02-15 16:16 ` [Qemu-devel] [PATCH 03/10] target/ppc: Fix support for "STOP light" states on POWER9 Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 04/10] target/ppc: Move "wakeup reset" code to a separate function Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 05/10] target/ppc: Rename "in_pm_state" to "resume_as_sreset" Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 06/10] target/ppc: Add POWER9 exception model Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 07/10] target/ppc: Detect erroneous condition in interrupt delivery Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 08/10] target/ppc: Add Hypervisor Virtualization Interrupt on POWER9 Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 09/10] target/ppc: Add POWER9 external interrupt model Cédric Le Goater
2019-02-15 16:16 ` [Qemu-devel] [PATCH 10/10] target/ppc: Add support for LPCR:HEIC on POWER9 Cédric Le Goater
2019-02-19  0:44 ` [Qemu-devel] [PATCH 00/10] ppc: Add Hypervisor Virtualization Interrupt " David Gibson

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=20190215161648.9600-3-clg@kaod.org \
    --to=clg@kaod.org \
    --cc=benh@kernel.crashing.org \
    --cc=david@gibson.dropbear.id.au \
    --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).