[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [RFC v3 11/19] tcg: add options for enabling MTTCG
From: |
Sergey Fedorov |
Subject: |
Re: [Qemu-devel] [RFC v3 11/19] tcg: add options for enabling MTTCG |
Date: |
Tue, 28 Jun 2016 00:07:11 +0300 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.8.0 |
On 03/06/16 23:40, Alex Bennée wrote:
> diff --git a/cpus.c b/cpus.c
> index 4cc2ce6..1694ce9 100644
> --- a/cpus.c
> +++ b/cpus.c
> @@ -25,6 +25,7 @@
> /* Needed early for CONFIG_BSD etc. */
> #include "qemu/osdep.h"
> #include "qemu-common.h"
> +#include "qemu/config-file.h"
> #include "cpu.h"
> #include "monitor/monitor.h"
> #include "qapi/qmp/qerror.h" @@ -148,6 +149,33 @@ typedef struct TimersState
> { } TimersState;
> static TimersState timers_state; +static bool mttcg_enabled; + +static
> bool default_mttcg_enabled(void) +{ + /* + * TODO: Check if we have a
> chance to have MTTCG working on this guest/host. + * Basically is the
> atomic instruction implemented? Is there any + * memory ordering
> issue? + */ + return false; +} + +void qemu_tcg_configure(QemuOpts
> *opts) +{ + const char *t = qemu_opt_get(opts, "thread");
> + if (t) {
> + mttcg_enabled = (strcmp(t, "multi") == 0);
> + } else {
> + mttcg_enabled = default_mttcg_enabled();
Wouldn't we like to check for user errors here? Suppose a user have
passed "multy" or "mult" and think they've enabled MTTCG whereas such
values are silently ignored.
> + }
> +}
> +
> +bool qemu_tcg_mttcg_enabled(void)
> +{
> + return mttcg_enabled;
> +}
> +
>
> int64_t cpu_get_icount_raw(void)
> {
(snip)
> diff --git a/qemu-options.hx b/qemu-options.hx
> index 6106520..dc865ec 100644
> --- a/qemu-options.hx
> +++ b/qemu-options.hx
> @@ -99,6 +99,26 @@ STEXI
> Select CPU model (@code{-cpu help} for list and additional feature selection)
> ETEXI
>
> +DEF("accel", HAS_ARG, QEMU_OPTION_accel,
> + "-accel [accel=]accelerator[,thread=single|multi]\n"
> + " select accelerator ('-accel help for list')\n"
> + " thread=single|multi (enable multi-threaded TCG)",
> QEMU_ARCH_ALL)
"mttcg=on|off" or "multithread=on|off", instead of
"thread=single|multi", are another possible variants. The former has
less letters :)
> +STEXI
> address@hidden -accel @var{name}[,address@hidden,...]]
> address@hidden -accel
> +This is used to enable an accelerator. Depending on the target architecture,
> +kvm, xen, or tcg can be available. By default, tcg is used. If there is more
> +than one accelerator specified, the next one is used if the previous one
> fails
> +to initialize.
> address@hidden @option
> address@hidden thread=single|multi
> +Controls number of TCG threads. When the TCG is multi-threaded there will be
> one
> +thread per vCPU therefor taking advantage of additional host cores. The
> default
> +is to enable multi-threading where both the back-end and front-ends support
> it and
> +no incompatible TCG features have been enabled (e.g. icount/replay).
We could raise and error on attempts to use icount with MTTCG enabled
modifying this piece of code in vl.c:
if (icount_opts) {
if (kvm_enabled() || xen_enabled()) {
error_report("-icount is not allowed with kvm or xen");
exit(1);
}
> address@hidden table
> +ETEXI
> +
> DEF("smp", HAS_ARG, QEMU_OPTION_smp,
> "-smp
> [cpus=]n[,maxcpus=cpus][,cores=cores][,threads=threads][,sockets=sockets]\n"
> " set the number of CPUs to 'n' [default=1]\n"
> diff --git a/vl.c b/vl.c
> index 18d1423..b1224f9 100644
> --- a/vl.c
> +++ b/vl.c
(snip)
> @@ -3682,6 +3704,27 @@ int main(int argc, char **argv, char **envp)
> qdev_prop_register_global(&kvm_pit_lost_tick_policy);
> break;
> }
> + case QEMU_OPTION_accel:
> + opts = qemu_opts_parse_noisily(qemu_find_opts("accel"),
> + optarg, true);
> + optarg = qemu_opt_get(opts, "accel");
> +
> + olist = qemu_find_opts("machine");
> + if (strcmp("kvm", optarg) == 0) {
> + qemu_opts_parse_noisily(olist, "accel=kvm", false);
> + } else if (strcmp("xen", optarg) == 0) {
> + qemu_opts_parse_noisily(olist, "accel=xen", false);
> + } else if (strcmp("tcg", optarg) == 0) {
> + qemu_opts_parse_noisily(olist, "accel=tcg", false);
> + qemu_tcg_configure(opts);
> + } else {
> + if (!is_help_option(optarg)) {
> + error_printf("Unknown accelerator: %s", optarg);
> + }
> + error_printf("Supported accelerators: kvm, xen, tcg\n");
> + exit(1);
> + }
I am wondering if we should use accel_find() here like in
configure_accelerator() and probably also make checks with
AccelClass::available().
Please consider wrapping this code in a separate function.
> + break;
> case QEMU_OPTION_usb:
> olist = qemu_find_opts("machine");
> qemu_opts_parse_noisily(olist, "usb=on", false);
Kind regards,
Sergey
- [Qemu-devel] [RFC v3 06/19] tcg: comment on which functions have to be called with tb_lock held, (continued)
- [Qemu-devel] [RFC v3 06/19] tcg: comment on which functions have to be called with tb_lock held, Alex Bennée, 2016/06/03
- [Qemu-devel] [RFC v3 04/19] docs: new design document multi-thread-tcg.txt (DRAFTING), Alex Bennée, 2016/06/03
- [Qemu-devel] [RFC v3 07/19] translate-all: Add assert_memory_lock annotations, Alex Bennée, 2016/06/03
- [Qemu-devel] [RFC v3 12/19] tcg: add kick timer for single-threaded vCPU emulation, Alex Bennée, 2016/06/03
- [Qemu-devel] [RFC v3 08/19] tcg: protect TBContext with tb_lock., Alex Bennée, 2016/06/03
- [Qemu-devel] [RFC v3 11/19] tcg: add options for enabling MTTCG, Alex Bennée, 2016/06/03
- Re: [Qemu-devel] [RFC v3 11/19] tcg: add options for enabling MTTCG,
Sergey Fedorov <=
- [Qemu-devel] [RFC v3 13/19] tcg: rename tcg_current_cpu to tcg_current_rr_cpu, Alex Bennée, 2016/06/03
Re: [Qemu-devel] [RFC v3 13/19] tcg: rename tcg_current_cpu to tcg_current_rr_cpu, Alex Bennée, 2016/06/07
[Qemu-devel] [RFC v3 14/19] tcg: remove global exit_request, Alex Bennée, 2016/06/03
[Qemu-devel] [RFC v3 09/19] target-arm/arm-powerctl: wake up sleeping CPUs, Alex Bennée, 2016/06/03