ppc/xive: Fix pulling pool and phys contexts

This improves the implementation of pulling pool and phys contexts in
XIVE1, by following closer the OS pulling code.

In particular, the old ring data is returned rather than the modified,
and irq signals are reset on pull.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Glenn Miles <milesg@linux.ibm.com>
Reviewed-by: Michael Kowal <kowal@linux.ibm.com>
Reviewed-by: Caleb Schlossin <calebs@linux.ibm.com>
Tested-by: Gautam Menghani <gautam@linux.ibm.com>
Link: https://lore.kernel.org/qemu-devel/20250512031100.439842-17-npiggin@gmail.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>

diff --git a/hw/intc/xive.c b/hw/intc/xive.c
index 8537cad27b2a395e83f9e9d34040ba98cd0d60d3..5483a815ef077b896638298eea620adb205e2b0e 100644 (file)
--- a/hw/intc/xive.c
+++ b/hw/intc/xive.c
@@ -241,25 +241,75 @@ static uint64_t xive_tm_ack_hv_reg(XivePresenter *xptr, XiveTCTX *tctx,
     return xive_tctx_accept(tctx, TM_QW3_HV_PHYS);
 }
 
+static void xive_pool_cam_decode(uint32_t cam, uint8_t *nvt_blk,
+                                 uint32_t *nvt_idx, bool *vp)
+{
+    if (nvt_blk) {
+        *nvt_blk = xive_nvt_blk(cam);
+    }
+    if (nvt_idx) {
+        *nvt_idx = xive_nvt_idx(cam);
+    }
+    if (vp) {
+        *vp = !!(cam & TM_QW2W2_VP);
+    }
+}
+
+static uint32_t xive_tctx_get_pool_cam(XiveTCTX *tctx, uint8_t *nvt_blk,
+                                       uint32_t *nvt_idx, bool *vp)
+{
+    uint32_t qw2w2 = xive_tctx_word2(&tctx->regs[TM_QW2_HV_POOL]);
+    uint32_t cam = be32_to_cpu(qw2w2);
+
+    xive_pool_cam_decode(cam, nvt_blk, nvt_idx, vp);
+    return qw2w2;
+}
+
+static void xive_tctx_set_pool_cam(XiveTCTX *tctx, uint32_t qw2w2)
+{
+    memcpy(&tctx->regs[TM_QW2_HV_POOL + TM_WORD2], &qw2w2, 4);
+}
+
 static uint64_t xive_tm_pull_pool_ctx(XivePresenter *xptr, XiveTCTX *tctx,
                                       hwaddr offset, unsigned size)
 {
-    uint32_t qw2w2_prev = xive_tctx_word2(&tctx->regs[TM_QW2_HV_POOL]);
     uint32_t qw2w2;
+    uint32_t qw2w2_new;
+    uint8_t nvt_blk;
+    uint32_t nvt_idx;
+    bool vp;
 
-    qw2w2 = xive_set_field32(TM_QW2W2_VP, qw2w2_prev, 0);
-    memcpy(&tctx->regs[TM_QW2_HV_POOL + TM_WORD2], &qw2w2, 4);
+    qw2w2 = xive_tctx_get_pool_cam(tctx, &nvt_blk, &nvt_idx, &vp);
+
+    if (!vp) {
+        qemu_log_mask(LOG_GUEST_ERROR, "XIVE: pull invalid POOL NVT %x/%x !?\n",
+                      nvt_blk, nvt_idx);
+    }
+
+    /* Invalidate CAM line */
+    qw2w2_new = xive_set_field32(TM_QW2W2_VP, qw2w2, 0);
+    xive_tctx_set_pool_cam(tctx, qw2w2_new);
+
+    xive_tctx_reset_signal(tctx, TM_QW1_OS);
+    xive_tctx_reset_signal(tctx, TM_QW2_HV_POOL);
     return qw2w2;
 }
 
 static uint64_t xive_tm_pull_phys_ctx(XivePresenter *xptr, XiveTCTX *tctx,
                                       hwaddr offset, unsigned size)
 {
-    uint8_t qw3b8_prev = tctx->regs[TM_QW3_HV_PHYS + TM_WORD2];
-    uint8_t qw3b8;
+    uint8_t qw3b8 = tctx->regs[TM_QW3_HV_PHYS + TM_WORD2];
+    uint8_t qw3b8_new;
+
+    qw3b8 = tctx->regs[TM_QW3_HV_PHYS + TM_WORD2];
+    if (!(qw3b8 & TM_QW3B8_VT)) {
+        qemu_log_mask(LOG_GUEST_ERROR, "XIVE: pulling invalid PHYS thread!?\n");
+    }
+    qw3b8_new = qw3b8 & ~TM_QW3B8_VT;
+    tctx->regs[TM_QW3_HV_PHYS + TM_WORD2] = qw3b8_new;
 
-    qw3b8 = qw3b8_prev & ~TM_QW3B8_VT;
-    tctx->regs[TM_QW3_HV_PHYS + TM_WORD2] = qw3b8;
+    xive_tctx_reset_signal(tctx, TM_QW1_OS);
+    xive_tctx_reset_signal(tctx, TM_QW3_HV_PHYS);
     return qw3b8;
 }
 
@@ -489,7 +539,7 @@ static uint64_t xive_tm_pull_os_ctx(XivePresenter *xptr, XiveTCTX *tctx,
     qw1w2 = xive_tctx_get_os_cam(tctx, &nvt_blk, &nvt_idx, &vo);
 
     if (!vo) {
-        qemu_log_mask(LOG_GUEST_ERROR, "XIVE: pulling invalid NVT %x/%x !?\n",
+        qemu_log_mask(LOG_GUEST_ERROR, "XIVE: pull invalid OS NVT %x/%x !?\n",
                       nvt_blk, nvt_idx);
     }