All of lore.kernel.org
 help / color / mirror / Atom feed
From: Michael Clark <mjc@sifive.com>
To: qemu-devel@nongnu.org
Cc: patches@groups.riscv.org, Michael Clark <mjc@sifive.com>,
	Palmer Dabbelt <palmer@sifive.com>,
	Sagar Karandikar <sagark@eecs.berkeley.edu>,
	Bastian Koppelmann <kbastian@mail.uni-paderborn.de>,
	Alistair Francis <Alistair.Francis@wdc.com>
Subject: [Qemu-devel] [PATCH v1 19/30] RISC-V: Allow interrupt controllers to claim interrupts
Date: Wed, 23 May 2018 12:15:06 +1200	[thread overview]
Message-ID: <1527034517-7851-20-git-send-email-mjc@sifive.com> (raw)
In-Reply-To: <1527034517-7851-1-git-send-email-mjc@sifive.com>

We can't allow the supervisor to control SEIP as this would allow the
supervisor to clear a pending external interrupt which will result in
lost a interrupt in the case a PLIC is attached. The SEIP bit must be
hardware controlled when a PLIC is attached.

This logic was previously hard-coded so SEIP was always masked even
if no PLIC was attached. This patch adds riscv_cpu_claim_interrupts
so that the PLIC can register control of SEIP. In the case of models
without a PLIC (spike), the SEIP bit remains software controlled.

This interface allows for hardware control of supervisor timer and
software interrupts by other interrupt controller models.

Cc: Palmer Dabbelt <palmer@sifive.com>
Cc: Sagar Karandikar <sagark@eecs.berkeley.edu>
Cc: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Cc: Alistair Francis <Alistair.Francis@wdc.com>
Signed-off-by: Michael Clark <mjc@sifive.com>
---
 hw/riscv/sifive_plic.c    | 13 +++++++++++++
 target/riscv/cpu.h        |  2 ++
 target/riscv/cpu_helper.c | 11 +++++++++++
 target/riscv/csr.c        | 12 ++++--------
 4 files changed, 30 insertions(+), 8 deletions(-)

diff --git a/hw/riscv/sifive_plic.c b/hw/riscv/sifive_plic.c
index dc6f4924e282..28e28d932f7c 100644
--- a/hw/riscv/sifive_plic.c
+++ b/hw/riscv/sifive_plic.c
@@ -23,6 +23,7 @@
 #include "qemu/error-report.h"
 #include "hw/sysbus.h"
 #include "target/riscv/cpu.h"
+#include "sysemu/sysemu.h"
 #include "hw/riscv/sifive_plic.h"
 
 #define RISCV_DEBUG_PLIC 0
@@ -447,6 +448,18 @@ static void sifive_plic_realize(DeviceState *dev, Error **errp)
     for (i = 0; i <= plic->num_sources; i++) {
         plic->irqs[i] = qemu_allocate_irq(sifive_plic_irq_request, plic, i);
     }
+
+    /* We can't allow the supervisor to control SEIP as this would allow the
+     * supervisor to clear a pending external interrupt which will result in
+     * lost a interrupt in the case a PLIC is attached. The SEIP bit must be
+     * hardware controlled when a PLIC is attached. */
+    for (i = 0; i < smp_cpus; i++) {
+        RISCVCPU *cpu = RISCV_CPU(qemu_get_cpu(i));
+        if (riscv_cpu_claim_interrupts(cpu, MIP_SEIP) < 0) {
+            error_report("sifive_plic_realize: SEIP already claimed");
+            exit(1);
+        }
+    }
 }
 
 static void sifive_plic_class_init(ObjectClass *klass, void *data)
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index d6bb3136db18..ae0e3f6a544d 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -138,6 +138,7 @@ struct CPURISCVState {
      * mip is 32-bits to allow atomic_read on 32-bit hosts.
      */
     uint32_t mip;
+    uint32_t miclaim;
 
     target_ulong mie;
     target_ulong mideleg;
@@ -262,6 +263,7 @@ void riscv_cpu_list(FILE *f, fprintf_function cpu_fprintf);
 #define cpu_mmu_index riscv_cpu_mmu_index
 
 #ifndef CONFIG_USER_ONLY
+int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint32_t interrupts);
 uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value);
 #define BOOL_TO_MASK(x) (-!!(x)) /* helper for riscv_cpu_update_mip value */
 #endif
diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
index b4bbf7a9fa0a..7c9f6c46c75a 100644
--- a/target/riscv/cpu_helper.c
+++ b/target/riscv/cpu_helper.c
@@ -72,6 +72,17 @@ bool riscv_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
 
 #if !defined(CONFIG_USER_ONLY)
 
+int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint32_t interrupts)
+{
+    CPURISCVState *env = &cpu->env;
+    if (env->miclaim & interrupts) {
+        return -1;
+    } else {
+        env->miclaim |= interrupts;
+        return 0;
+    }
+}
+
 /* iothread_mutex must be held */
 uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value)
 {
diff --git a/target/riscv/csr.c b/target/riscv/csr.c
index 45e33d876034..9bbe81a110a5 100644
--- a/target/riscv/csr.c
+++ b/target/riscv/csr.c
@@ -502,15 +502,11 @@ static int rmw_mip(CPURISCVState *env, int csrno, target_ulong *ret_value,
                    target_ulong new_value, target_ulong write_mask)
 {
     RISCVCPU *cpu = riscv_env_get_cpu(env);
-    target_ulong mask = write_mask & delegable_ints;
-    uint32_t old_mip;
 
-    /* We can't allow the supervisor to control SEIP as this would allow the
-     * supervisor to clear a pending external interrupt which will result in
-     * lost a interrupt in the case a PLIC is attached. The SEIP bit must be
-     * hardware controlled when a PLIC is attached. This should be an option
-     * for CPUs with software-delegated Supervisor External Interrupts. */
-    mask &= ~MIP_SEIP;
+    /* Allow software control of delegable interrupts not claimed by hardware */
+    target_ulong mask = write_mask & delegable_ints & ~env->miclaim;
+
+    uint32_t old_mip;
 
     if (mask) {
         qemu_mutex_lock_iothread();
-- 
2.7.0

  parent reply	other threads:[~2018-05-23  0:18 UTC|newest]

Thread overview: 61+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-23  0:14 [Qemu-devel] [PATCH v1 00/30] QEMU 2.13 RISC-V updates Michael Clark
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 01/30] RISC-V: Update address bits to support sv39 and sv48 Michael Clark
2018-05-25 15:07   ` Richard Henderson
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 02/30] RISC-V: Improve page table walker spec compliance Michael Clark
2018-05-23 22:31   ` Michael Clark
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 03/30] RISC-V: Use atomic_cmpxchg to update PLIC bitmaps Michael Clark
2018-05-29 23:32   ` Alistair Francis
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 04/30] RISC-V: Simplify riscv_cpu_local_irqs_pending Michael Clark
2018-05-25 15:15   ` Richard Henderson
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 05/30] RISC-V: Allow setting and clearing multiple irqs Michael Clark
2018-05-23 23:55   ` Alistair Francis
2018-05-25 15:19   ` Richard Henderson
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 06/30] RISC-V: Move non-ops from op_helper to cpu_helper Michael Clark
2018-05-23 12:23   ` Philippe Mathieu-Daudé
2018-05-25 15:20   ` Richard Henderson
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 07/30] RISC-V: Update CSR and interrupt definitions Michael Clark
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 08/30] RISC-V: Implement modular CSR helper interface Michael Clark
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 09/30] RISC-V: Implement atomic mip/sip CSR updates Michael Clark
2018-05-29 23:34   ` Alistair Francis
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 10/30] RISC-V: Implement existential predicates for CSRs Michael Clark
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 11/30] RISC-V: Split out mstatus_fs from tb_flags Michael Clark
2018-05-23 12:25   ` Philippe Mathieu-Daudé
2018-05-29 23:40   ` Alistair Francis
2018-05-23  0:14 ` [Qemu-devel] [PATCH v1 12/30] RISC-V: Mark mstatus.fs dirty Michael Clark
2018-05-29 23:38   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 13/30] RISC-V: Implement mstatus.TSR/TW/TVM Michael Clark
2018-05-23 12:26   ` Philippe Mathieu-Daudé
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 14/30] RISC-V: Add public API for the CSR dispatch table Michael Clark
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 15/30] RISC-V: Add hartid and \n to interrupt logging Michael Clark
2018-05-23 12:33   ` Philippe Mathieu-Daudé
2018-05-24 22:47     ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 16/30] RISC-V: Use riscv prefix consistently on cpu helpers Michael Clark
2018-05-23 12:36   ` Philippe Mathieu-Daudé
2018-05-29 23:43   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 17/30] RISC-V: Replace __builtin_popcount with ctpop8 in PLIC Michael Clark
2018-05-23 12:37   ` Philippe Mathieu-Daudé
2018-05-29 23:47   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 18/30] RISC-V: Add missing free for plic_hart_config Michael Clark
2018-05-23 12:40   ` Philippe Mathieu-Daudé
2018-05-24 22:43     ` Alistair Francis
2018-05-23  0:15 ` Michael Clark [this message]
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 20/30] RISC-V: Add misa to DisasContext Michael Clark
2018-05-23 12:42   ` Philippe Mathieu-Daudé
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 21/30] RISC-V: Add misa.MAFD checks to translate Michael Clark
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 22/30] RISC-V: Add misa runtime write support Michael Clark
2018-05-25 18:53   ` Richard Henderson
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 23/30] RISC-V: Fix CLINT timecmp low 32-bit writes Michael Clark
2018-05-25 22:40   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 24/30] RISC-V: Fix PLIC pending bitfield reads Michael Clark
2018-05-25 22:38   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 25/30] RISC-V: Enable second UART on sifive_e and sifive_u Michael Clark
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 26/30] RISC-V: Remove unnecessary disassembler constraints Michael Clark
2018-05-24 22:45   ` Alistair Francis
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 27/30] elf: Add RISC-V PSABI ELF header defines Michael Clark
2018-05-23  6:44   ` Laurent Vivier
2018-05-25  7:17     ` Michael Clark
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 28/30] RISC-V: linux-user support for RVE ABI Michael Clark
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 29/30] RISC-V: Don't add NULL bootargs to device-tree Michael Clark
2018-05-23 12:45   ` Philippe Mathieu-Daudé
2018-05-23  0:15 ` [Qemu-devel] [PATCH v1 30/30] RISC-V: Support separate firmware and kernel payload Michael Clark
2018-05-23 12:49   ` Philippe Mathieu-Daudé

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=1527034517-7851-20-git-send-email-mjc@sifive.com \
    --to=mjc@sifive.com \
    --cc=Alistair.Francis@wdc.com \
    --cc=kbastian@mail.uni-paderborn.de \
    --cc=palmer@sifive.com \
    --cc=patches@groups.riscv.org \
    --cc=qemu-devel@nongnu.org \
    --cc=sagark@eecs.berkeley.edu \
    /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.