From: Stefan Weil <sw@weilnetz.de>
To: Richard Henderson <richard.henderson@linaro.org>, qemu-devel@nongnu.org
Subject: Increased execution time with TCI in latest git master (was: Re: [PULL 00/46] tcg patch queue)
Date: Sat, 6 Feb 2021 20:38:49 +0100 [thread overview]
Message-ID: <02ad8dba-48a5-b547-05ba-d7fa1150b1d0@weilnetz.de> (raw)
In-Reply-To: <20210205225650.1330794-1-richard.henderson@linaro.org>
Am 05.02.21 um 23:56 schrieb Richard Henderson:
> The following changes since commit d0dddab40e472ba62b5f43f11cc7dba085dabe71:
>
> Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging (2021-02-05 15:27:02 +0000)
>
> are available in the Git repository at:
>
> https://gitlab.com/rth7680/qemu.git tags/pull-tcg-20210205
>
> for you to fetch changes up to fb6916dd6ca8bb4b42d44baba9c67ecaf2279577:
>
> accel: introduce AccelCPUClass extending CPUClass (2021-02-05 10:24:15 -1000)
>
> ----------------------------------------------------------------
> TCGCPUOps cleanups (claudio)
> tcg/s390 compare fix (phil)
> tcg/aarch64 rotli_vec fix
> tcg/tci cleanups and fixes
Hi Richard,
I tested pull-tcg-20210205 on a sparc64 machine.
Timing results for BIOS booting ("./qemu-system-x86_64", no image):
Before pull-tcg-20210205 with my latest patches for TCI:
TCG native: 8 s user time, TCG interpreter: 40 s user time
With pull-tcg-20210205 and
https://patchwork.kernel.org/project/qemu-devel/patch/20210127123446.2037721-1-sw@weilnetz.de/:
TCG native: 8 s user time, TCG interpreter: 50 s user time
So for this test case TCI was about 5 times slower than native TCG, but
the latest changes increased the execution time by about 25 %.
I am still searching what caused this detoriation. My first suspect was
thread local storage, but that wasn't it. Do you have any idea?
Stefan
next prev parent reply other threads:[~2021-02-06 19:39 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-02-05 22:56 [PULL 00/46] tcg patch queue Richard Henderson
2021-02-05 22:56 ` [PULL 01/46] tcg/s390: Fix compare instruction from extended-immediate facility Richard Henderson
2021-02-05 22:56 ` [PULL 02/46] exec/cpu-defs: Remove TCG backends dependency Richard Henderson
2021-02-05 22:56 ` [PULL 03/46] tcg/aarch64: Do not convert TCGArg to temps that are not temps Richard Henderson
2021-02-05 22:56 ` [PULL 04/46] configure: Fix --enable-tcg-interpreter Richard Henderson
2021-02-05 22:56 ` [PULL 05/46] tcg/tci: Make tci_tb_ptr thread-local Richard Henderson
2021-02-05 22:56 ` [PULL 06/46] tcg/tci: Implement INDEX_op_ld16s_i32 Richard Henderson
2021-02-05 22:56 ` [PULL 07/46] tcg/tci: Implement INDEX_op_ld8s_i64 Richard Henderson
2021-02-05 22:56 ` [PULL 08/46] tcg/tci: Inline tci_write_reg32s into the only caller Richard Henderson
2021-02-05 22:56 ` [PULL 09/46] tcg/tci: Inline tci_write_reg8 into its callers Richard Henderson
2021-02-05 22:56 ` [PULL 10/46] tcg/tci: Inline tci_write_reg16 into the only caller Richard Henderson
2021-02-05 22:56 ` [PULL 11/46] tcg/tci: Inline tci_write_reg32 into all callers Richard Henderson
2021-02-05 22:56 ` [PULL 12/46] tcg/tci: Inline tci_write_reg64 into 64-bit callers Richard Henderson
2021-02-05 22:56 ` [PULL 13/46] tcg/tci: Merge INDEX_op_ld8u_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 14/46] tcg/tci: Merge INDEX_op_ld8s_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 15/46] tcg/tci: Merge INDEX_op_ld16u_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 16/46] tcg/tci: Merge INDEX_op_ld16s_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 17/46] tcg/tci: Merge INDEX_op_{ld_i32,ld32u_i64} Richard Henderson
2021-02-05 22:56 ` [PULL 18/46] tcg/tci: Merge INDEX_op_st8_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 19/46] tcg/tci: Merge INDEX_op_st16_{i32,i64} Richard Henderson
2021-02-05 22:56 ` [PULL 20/46] tcg/tci: Move stack bounds check to compile-time Richard Henderson
2021-02-05 22:56 ` [PULL 21/46] tcg/tci: Merge INDEX_op_{st_i32,st32_i64} Richard Henderson
2021-02-05 22:56 ` [PULL 22/46] tcg/tci: Use g_assert_not_reached Richard Henderson
2021-02-05 22:56 ` [PULL 23/46] tcg/tci: Remove dead code for TCG_TARGET_HAS_div2_* Richard Henderson
2021-02-05 22:56 ` [PULL 24/46] tcg/tci: Implement 64-bit division Richard Henderson
2021-02-05 22:56 ` [PULL 25/46] tcg/tci: Remove TODO as unused Richard Henderson
2021-02-05 22:56 ` [PULL 26/46] tcg/tci: Restrict TCG_TARGET_NB_REGS to 16 Richard Henderson
2021-02-05 22:56 ` [PULL 27/46] tcg/tci: Fix TCG_REG_R4 misusage Richard Henderson
2021-02-05 22:56 ` [PULL 28/46] tcg/tci: Remove TCG_CONST Richard Henderson
2021-02-05 22:56 ` [PULL 29/46] cpu: Introduce TCGCpuOperations struct Richard Henderson
2021-02-05 22:56 ` [PULL 30/46] target/riscv: remove CONFIG_TCG, as it is always TCG Richard Henderson
2021-02-05 22:56 ` [PULL 31/46] accel/tcg: split TCG-only code from cpu_exec_realizefn Richard Henderson
2021-02-05 22:56 ` [PULL 32/46] cpu: Move synchronize_from_tb() to tcg_ops Richard Henderson
2021-02-05 22:56 ` [PULL 33/46] cpu: Move cpu_exec_* " Richard Henderson
2021-02-05 22:56 ` [PULL 34/46] cpu: Move tlb_fill " Richard Henderson
2021-02-05 22:56 ` [PULL 35/46] cpu: Move debug_excp_handler " Richard Henderson
2021-02-05 22:56 ` [PULL 36/46] target/arm: do not use cc->do_interrupt for KVM directly Richard Henderson
2021-02-05 22:56 ` [PULL 37/46] cpu: move cc->do_interrupt to tcg_ops Richard Henderson
2021-02-05 22:56 ` [PULL 38/46] cpu: move cc->transaction_failed " Richard Henderson
2021-02-23 21:43 ` Philippe Mathieu-Daudé
2021-02-24 8:46 ` Claudio Fontana
2021-02-05 22:56 ` [PULL 39/46] cpu: move do_unaligned_access " Richard Henderson
2021-02-05 22:56 ` [PULL 40/46] physmem: make watchpoint checking code TCG-only Richard Henderson
2021-02-05 22:56 ` [PULL 41/46] cpu: move adjust_watchpoint_address to tcg_ops Richard Henderson
2021-02-05 22:56 ` [PULL 42/46] cpu: move debug_check_watchpoint " Richard Henderson
2021-02-05 22:56 ` [PULL 43/46] cpu: tcg_ops: move to tcg-cpu-ops.h, keep a pointer in CPUClass Richard Henderson
2021-02-05 22:56 ` [PULL 44/46] accel: extend AccelState and AccelClass to user-mode Richard Henderson
2021-02-05 22:56 ` [PULL 45/46] accel: replace struct CpusAccel with AccelOpsClass Richard Henderson
2021-02-05 22:56 ` [PULL 46/46] accel: introduce AccelCPUClass extending CPUClass Richard Henderson
2021-04-26 14:42 ` Philippe Mathieu-Daudé
2021-02-06 14:28 ` [PULL 00/46] tcg patch queue Peter Maydell
2021-02-06 19:14 ` Philippe Mathieu-Daudé
2021-02-06 19:38 ` Stefan Weil [this message]
2021-02-07 3:45 ` Increased execution time with TCI in latest git master (was: Re: [PULL 00/46] tcg patch queue) Richard Henderson
2021-02-07 10:50 ` Stefan Weil
2021-02-07 18:37 ` Richard Henderson
2021-02-07 22:00 ` Stefan Weil
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=02ad8dba-48a5-b547-05ba-d7fa1150b1d0@weilnetz.de \
--to=sw@weilnetz.de \
--cc=qemu-devel@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).