qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: David Gibson <david@gibson.dropbear.id.au>
To: peter.maydell@linaro.org
Cc: Richard Henderson <richard.henderson@linaro.org>,
	mark.cave-ayland@ilande.co.uk, qemu-devel@nongnu.org,
	groug@kaod.org, Luis Pires <luis.pires@eldorado.org.br>,
	hpoussin@reactos.org, clg@kaod.org, qemu-ppc@nongnu.org,
	philmd@redhat.com, David Gibson <david@gibson.dropbear.id.au>
Subject: [PULL 13/44] target/ppc: fix setting of CR flags in bcdcfsq
Date: Thu, 30 Sep 2021 15:43:55 +1000	[thread overview]
Message-ID: <20210930054426.357344-14-david@gibson.dropbear.id.au> (raw)
In-Reply-To: <20210930054426.357344-1-david@gibson.dropbear.id.au>

From: Luis Pires <luis.pires@eldorado.org.br>

According to the ISA, CR should be set based on the source value, and
not on the packed decimal result.
The way this was implemented would cause GT, LT and EQ to be set
incorrectly when the source value was too large and the 31 least
significant digits of the packed decimal result ended up being all zero.
This would happen for source values of +/-10^31, +/-10^32, etc.

The new implementation fixes this and also skips the result calculation
altogether in case of src overflow.

Signed-off-by: Luis Pires <luis.pires@eldorado.org.br>
Message-Id: <20210823150235.35759-1-luis.pires@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 target/ppc/int_helper.c | 61 ++++++++++++++++++++++++++++++++---------
 1 file changed, 48 insertions(+), 13 deletions(-)

diff --git a/target/ppc/int_helper.c b/target/ppc/int_helper.c
index c2d3248d1e..f5dac3aa87 100644
--- a/target/ppc/int_helper.c
+++ b/target/ppc/int_helper.c
@@ -2480,10 +2480,26 @@ uint32_t helper_bcdctz(ppc_avr_t *r, ppc_avr_t *b, uint32_t ps)
     return cr;
 }
 
+/**
+ * Compare 2 128-bit unsigned integers, passed in as unsigned 64-bit pairs
+ *
+ * Returns:
+ * > 0 if ahi|alo > bhi|blo,
+ * 0 if ahi|alo == bhi|blo,
+ * < 0 if ahi|alo < bhi|blo
+ */
+static inline int ucmp128(uint64_t alo, uint64_t ahi,
+                          uint64_t blo, uint64_t bhi)
+{
+    return (ahi == bhi) ?
+        (alo > blo ? 1 : (alo == blo ? 0 : -1)) :
+        (ahi > bhi ? 1 : -1);
+}
+
 uint32_t helper_bcdcfsq(ppc_avr_t *r, ppc_avr_t *b, uint32_t ps)
 {
     int i;
-    int cr = 0;
+    int cr;
     uint64_t lo_value;
     uint64_t hi_value;
     ppc_avr_t ret = { .u64 = { 0, 0 } };
@@ -2492,28 +2508,47 @@ uint32_t helper_bcdcfsq(ppc_avr_t *r, ppc_avr_t *b, uint32_t ps)
         lo_value = -b->VsrSD(1);
         hi_value = ~b->VsrD(0) + !lo_value;
         bcd_put_digit(&ret, 0xD, 0);
+
+        cr = CRF_LT;
     } else {
         lo_value = b->VsrD(1);
         hi_value = b->VsrD(0);
         bcd_put_digit(&ret, bcd_preferred_sgn(0, ps), 0);
-    }
 
-    if (divu128(&lo_value, &hi_value, 1000000000000000ULL) ||
-            lo_value > 9999999999999999ULL) {
-        cr = CRF_SO;
+        if (hi_value == 0 && lo_value == 0) {
+            cr = CRF_EQ;
+        } else {
+            cr = CRF_GT;
+        }
     }
 
-    for (i = 1; i < 16; hi_value /= 10, i++) {
-        bcd_put_digit(&ret, hi_value % 10, i);
-    }
+    /*
+     * Check src limits: abs(src) <= 10^31 - 1
+     *
+     * 10^31 - 1 = 0x0000007e37be2022 c0914b267fffffff
+     */
+    if (ucmp128(lo_value, hi_value,
+                0xc0914b267fffffffULL, 0x7e37be2022ULL) > 0) {
+        cr |= CRF_SO;
 
-    for (; i < 32; lo_value /= 10, i++) {
-        bcd_put_digit(&ret, lo_value % 10, i);
-    }
+        /*
+         * According to the ISA, if src wouldn't fit in the destination
+         * register, the result is undefined.
+         * In that case, we leave r unchanged.
+         */
+    } else {
+        divu128(&lo_value, &hi_value, 1000000000000000ULL);
 
-    cr |= bcd_cmp_zero(&ret);
+        for (i = 1; i < 16; hi_value /= 10, i++) {
+            bcd_put_digit(&ret, hi_value % 10, i);
+        }
 
-    *r = ret;
+        for (; i < 32; lo_value /= 10, i++) {
+            bcd_put_digit(&ret, lo_value % 10, i);
+        }
+
+        *r = ret;
+    }
 
     return cr;
 }
-- 
2.31.1



  parent reply	other threads:[~2021-09-30  5:57 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-30  5:43 [PULL 00/44] ppc-for-6.2 queue 20210930 David Gibson
2021-09-30  5:43 ` [PULL 01/44] host-utils: Fix overflow detection in divu128() David Gibson
2021-09-30  5:43 ` [PULL 02/44] host-utils: fix missing zero-extension in divs128 David Gibson
2021-09-30  5:43 ` [PULL 03/44] host-utils: introduce uabs64() David Gibson
2021-09-30  5:43 ` [PULL 04/44] i386/kvm: Replace abs64() with uabs64() from host-utils David Gibson
2021-09-30  5:43 ` [PULL 05/44] ppc/spapr: Add a POWER10 DD2 CPU David Gibson
2021-09-30  5:43 ` [PULL 06/44] ppc/pnv: Add a comment on the "primary-topology-index" property David Gibson
2021-09-30  5:43 ` [PULL 07/44] ppc/pnv: Remove useless variable David Gibson
2021-09-30  5:43 ` [PULL 08/44] ppc/xive: Export priority_to_ipb() helper David Gibson
2021-09-30  5:43 ` [PULL 09/44] ppc/xive: Export xive_tctx_word2() helper David Gibson
2021-09-30  5:43 ` [PULL 10/44] ppc/pnv: Rename "id" to "quad-id" in PnvQuad David Gibson
2021-09-30  5:43 ` [PULL 11/44] docs/system: ppc: Update the URL for OpenPOWER firmware images David Gibson
2021-09-30  5:43 ` [PULL 12/44] ppc/pnv: Add an assert when calculating the RAM distribution on chips David Gibson
2021-09-30  5:43 ` David Gibson [this message]
2021-09-30  5:43 ` [PULL 14/44] memory_hotplug.c: handle dev->id = NULL in acpi_memory_hotplug_write() David Gibson
2021-09-30  5:43 ` [PULL 15/44] spapr.c: handle dev->id in spapr_memory_unplug_rollback() David Gibson
2021-09-30  5:43 ` [PULL 16/44] spapr_drc.c: do not error_report() when drc->dev->id == NULL David Gibson
2021-09-30  5:43 ` [PULL 17/44] qapi/qdev.json: fix DEVICE_DELETED parameters doc David Gibson
2021-09-30  5:44 ` [PULL 18/44] qapi/qdev.json: add DEVICE_UNPLUG_GUEST_ERROR QAPI event David Gibson
2021-09-30  5:44 ` [PULL 19/44] spapr: use DEVICE_UNPLUG_GUEST_ERROR to report unplug errors David Gibson
2021-09-30  5:44 ` [PULL 20/44] memory_hotplug.c: send DEVICE_UNPLUG_GUEST_ERROR in acpi_memory_hotplug_write() David Gibson
2021-09-30  5:44 ` [PULL 21/44] target/ppc: Convert debug to trace events (exceptions) David Gibson
2021-09-30  5:44 ` [PULL 22/44] target/ppc: Replace debug messages by asserts for unknown IRQ pins David Gibson
2021-09-30  5:44 ` [PULL 23/44] target/ppc: add LPCR[HR] to DisasContext and hflags David Gibson
2021-09-30  5:44 ` [PULL 24/44] target/ppc: Check privilege level based on PSR and LPCR[HR] in tlbie[l] David Gibson
2021-09-30  5:44 ` [PULL 25/44] spapr_numa.c: split FORM1 code into helpers David Gibson
2021-09-30  5:44 ` [PULL 26/44] spapr_numa.c: scrap 'legacy_numa' concept David Gibson
2021-09-30  5:44 ` [PULL 27/44] spapr_numa.c: parametrize FORM1 macros David Gibson
2021-09-30  5:44 ` [PULL 28/44] spapr_numa.c: rename numa_assoc_array to FORM1_assoc_array David Gibson
2021-09-30  5:44 ` [PULL 29/44] spapr: move FORM1 verifications to post CAS David Gibson
2021-09-30  5:44 ` [PULL 30/44] spapr_numa.c: FORM2 NUMA affinity support David Gibson
2021-09-30  5:44 ` [PULL 31/44] spapr_numa.c: handle auto NUMA node with no distance info David Gibson
2021-09-30  5:44 ` [PULL 32/44] target/ppc: Convert debug to trace events (decrementer and IRQ) David Gibson
2021-09-30  5:44 ` [PULL 33/44] target/ppc: Fix 64-bit decrementer David Gibson
2021-10-02 10:39   ` Peter Maydell
2021-10-04  6:54     ` Cédric Le Goater
2021-09-30  5:44 ` [PULL 34/44] hw/intc: openpic: Correct the reset value of IPIDR for FSL chipset David Gibson
2021-09-30  5:44 ` [PULL 35/44] hw/intc: openpic: Drop Raven related codes David Gibson
2021-09-30  5:44 ` [PULL 36/44] hw/intc: openpic: Clean up the styles David Gibson
2021-09-30  5:44 ` [PULL 37/44] spapr_numa.c: fixes in spapr_numa_FORM2_write_rtas_tables() David Gibson
2021-09-30  5:44 ` [PULL 38/44] spapr/xive: Fix kvm_xive_source_reset trace event David Gibson
2021-09-30  5:44 ` [PULL 39/44] MAINTAINERS: Remove machine specific files from ppc TCG CPUs entry David Gibson
2021-09-30  5:44 ` [PULL 40/44] MAINTAINERS: Remove David & Greg as reviewers for a number of boards David Gibson
2021-09-30  5:44 ` [PULL 41/44] MAINTAINERS: Orphan obscure ppc platforms David Gibson
2021-09-30  5:44 ` [PULL 42/44] MAINTAINERS: Remove David & Greg as reviewers/co-maintainers of powernv David Gibson
2021-09-30  5:44 ` [PULL 43/44] MAINTAINERS: Add information for OpenPIC David Gibson
2021-09-30  5:44 ` [PULL 44/44] MAINTAINERS: Demote sPAPR from "Supported" to "Maintained" David Gibson
2021-09-30 16:37 ` [PULL 00/44] ppc-for-6.2 queue 20210930 Peter Maydell

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=20210930054426.357344-14-david@gibson.dropbear.id.au \
    --to=david@gibson.dropbear.id.au \
    --cc=clg@kaod.org \
    --cc=groug@kaod.org \
    --cc=hpoussin@reactos.org \
    --cc=luis.pires@eldorado.org.br \
    --cc=mark.cave-ayland@ilande.co.uk \
    --cc=peter.maydell@linaro.org \
    --cc=philmd@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-ppc@nongnu.org \
    --cc=richard.henderson@linaro.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).