qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 0/2] target/arm: Fix sve2p1 mtedesc assertion
@ 2025-07-22 14:23 Richard Henderson
  2025-07-22 14:23 ` [PATCH 1/2] target/arm: Expand the descriptor for SME/SVE memory ops to i64 Richard Henderson
                   ` (3 more replies)
  0 siblings, 4 replies; 8+ messages in thread
From: Richard Henderson @ 2025-07-22 14:23 UTC (permalink / raw)
  To: qemu-devel; +Cc: qemu-arm, peter.maydell

Hi Peter,

Sadly, I couldn't reorg mtedesc as I hypothesized, because of
different usage within AdvSIMD.  So I decided to expand the
mtedesc from 17 to 32 bits, and then pack the gvec desc and
mtedesc into a 64-bit argument.

Lightly tested so far, but it does fix the LD3Q/LD4Q assert.


r~


Richard Henderson (2):
  target/arm: Expand the descriptor for SME/SVE memory ops to i64
  target/arm: Pack mtedesc into upper 32 bits of descriptor

 target/arm/internals.h         |    8 +-
 target/arm/tcg/helper-sme.h    |  144 ++--
 target/arm/tcg/helper-sve.h    | 1196 ++++++++++++++++----------------
 target/arm/tcg/translate-a64.h |    2 +-
 target/arm/tcg/sme_helper.c    |   30 +-
 target/arm/tcg/sve_helper.c    |  145 ++--
 target/arm/tcg/translate-sme.c |    6 +-
 target/arm/tcg/translate-sve.c |   38 +-
 8 files changed, 772 insertions(+), 797 deletions(-)

-- 
2.43.0



^ permalink raw reply	[flat|nested] 8+ messages in thread
* [PULL 0/1] tcg patch queue
@ 2025-07-21 15:11 Richard Henderson
  2025-07-21 15:11 ` [PULL 1/1] tcg/optimize: Don't fold INDEX_op_and_vec to extract Richard Henderson
  0 siblings, 1 reply; 8+ messages in thread
From: Richard Henderson @ 2025-07-21 15:11 UTC (permalink / raw)
  To: qemu-devel

The following changes since commit 56a3033abcfcf72a2f4f1376a605a0b1ad526b67:

  Merge tag 'pull-request-2025-07-21' of https://gitlab.com/thuth/qemu into staging (2025-07-21 06:34:56 -0400)

are available in the Git repository at:

  https://gitlab.com/rth7680/qemu.git tags/pull-tcg-20250721

for you to fetch changes up to 2c047bdb77e8f636936edd9ac5000521c9580477:

  tcg/optimize: Don't fold INDEX_op_and_vec to extract (2025-07-21 08:09:04 -0700)

----------------------------------------------------------------
tcg/optimize: Don't fold INDEX_op_and_vec to extract

----------------------------------------------------------------
Richard Henderson (1):
      tcg/optimize: Don't fold INDEX_op_and_vec to extract

 tcg/optimize.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)


^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2025-07-22 16:09 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-07-22 14:23 [PATCH 0/2] target/arm: Fix sve2p1 mtedesc assertion Richard Henderson
2025-07-22 14:23 ` [PATCH 1/2] target/arm: Expand the descriptor for SME/SVE memory ops to i64 Richard Henderson
2025-07-22 14:23 ` [PULL 1/1] tcg/optimize: Don't fold INDEX_op_and_vec to extract Richard Henderson
2025-07-22 15:22   ` Richard Henderson
2025-07-22 14:23 ` [PATCH 2/2] target/arm: Pack mtedesc into upper 32 bits of descriptor Richard Henderson
2025-07-22 14:52 ` [PATCH 0/2] target/arm: Fix sve2p1 mtedesc assertion Peter Maydell
2025-07-22 15:18   ` Richard Henderson
  -- strict thread matches above, loose matches on Subject: below --
2025-07-21 15:11 [PULL 0/1] tcg patch queue Richard Henderson
2025-07-21 15:11 ` [PULL 1/1] tcg/optimize: Don't fold INDEX_op_and_vec to extract Richard Henderson

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).