qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Michael Rolnik <mrolnik@gmail.com>
To: Aleksandar Markovic <aleksandar.m.mail@gmail.com>
Cc: "thuth@redhat.com" <thuth@redhat.com>,
	"richard.henderson@linaro.org" <richard.henderson@linaro.org>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"dovgaluk@ispras.ru" <dovgaluk@ispras.ru>,
	"imammedo@redhat.com" <imammedo@redhat.com>,
	"philmd@redhat.com" <philmd@redhat.com>
Subject: Re: [PATCH v32 04/13] target/avr: Add instruction translation - Registers definition
Date: Thu, 17 Oct 2019 23:43:04 +0300	[thread overview]
Message-ID: <CAK4993jaqFrzcmAyz7neJoq5cQgnB_o35dZhdh9_uVyOfKGG-Q@mail.gmail.com> (raw)
In-Reply-To: <CAL1e-=hGJQqVnvAu=ZJRV-AdoAWpkSvdf5ex=b7EwkodUGKpUg@mail.gmail.com>

On Thu, Oct 17, 2019 at 11:17 PM Aleksandar Markovic
<aleksandar.m.mail@gmail.com> wrote:
>>
>>
>> >> +static TCGv cpu_Cf;
>> >> +static TCGv cpu_Zf;
>> >> +static TCGv cpu_Nf;
>> >> +static TCGv cpu_Vf;
>> >> +static TCGv cpu_Sf;
>> >> +static TCGv cpu_Hf;
>> >> +static TCGv cpu_Tf;
>> >> +static TCGv cpu_If;
>> >> +
>> >
>> >
>> > Hello, Michael,
>> >
>> > Is there any particular reason or motivation beyond modelling status register flags as TCGv variables?
>>
>>
>>
>> I think it's easier this way as I don't need to convert flag values to
>> bits or bits to flag values.
>
>
> Ok. But, how do you map 0/1 flag value to the value of a TCGv variable and vice versa? In other words, what value or values (out of 2^32 vales) of a TCGv variable mean the flag is 1? And the same question for 0.
>
> Is 0110000111000010100 one or zero?
>
> Besides, in such arrangement, how do you display the 8-bit status register in gdb, if at all?

each flag register is either 0 or 1, they are calculated here
1. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L146-L148
2. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L166
3. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L185-L187
4. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L205
5. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L214-L215
6. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/translate.c#L222-L223
The COU itself never uses SREG at all, only the flags.

As for the GDB it's get assembled/disassembled here
1. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/cpu.h#L219-L243
2. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/gdbstub.c#L35-L37
3. https://github.com/michaelrolnik/qemu-avr/blob/avr-v32/target/avr/gdbstub.c#L66-L68

>
> A.
>
>>
>> >
>> > A.
>> >
>> >
>> >
>> >>
>> >> +static TCGv cpu_rampD;
>> >> +static TCGv cpu_rampX;
>> >> +static TCGv cpu_rampY;
>> >> +static TCGv cpu_rampZ;
>> >> +
>> >> +static TCGv cpu_r[NO_CPU_REGISTERS];
>> >> +static TCGv cpu_eind;
>> >> +static TCGv cpu_sp;
>> >> +
>> >> +static TCGv cpu_skip;
>> >> +
>> >> +static const char reg_names[NO_CPU_REGISTERS][8] = {
>> >> +    "r0",  "r1",  "r2",  "r3",  "r4",  "r5",  "r6",  "r7",
>> >> +    "r8",  "r9",  "r10", "r11", "r12", "r13", "r14", "r15",
>> >> +    "r16", "r17", "r18", "r19", "r20", "r21", "r22", "r23",
>> >> +    "r24", "r25", "r26", "r27", "r28", "r29", "r30", "r31",
>> >> +};
>> >> +#define REG(x) (cpu_r[x])
>> >> +
>> >> +enum {
>> >> +    DISAS_EXIT   = DISAS_TARGET_0,  /* We want return to the cpu main loop.  */
>> >> +    DISAS_LOOKUP = DISAS_TARGET_1,  /* We have a variable condition exit.  */
>> >> +    DISAS_CHAIN  = DISAS_TARGET_2,  /* We have a single condition exit.  */
>> >> +};
>> >> +
>> >> +typedef struct DisasContext DisasContext;
>> >> +
>> >> +/* This is the state at translation time. */
>> >> +struct DisasContext {
>> >> +    TranslationBlock *tb;
>> >> +
>> >> +    CPUAVRState *env;
>> >> +    CPUState *cs;
>> >> +
>> >> +    target_long npc;
>> >> +    uint32_t opcode;
>> >> +
>> >> +    /* Routine used to access memory */
>> >> +    int memidx;
>> >> +    int bstate;
>> >> +    int singlestep;
>> >> +
>> >> +    TCGv skip_var0;
>> >> +    TCGv skip_var1;
>> >> +    TCGCond skip_cond;
>> >> +    bool free_skip_var0;
>> >> +};
>> >> +
>> >> +static int to_A(DisasContext *ctx, int indx) { return 16 + (indx % 16); }
>> >> +static int to_B(DisasContext *ctx, int indx) { return 16 + (indx % 8); }
>> >> +static int to_C(DisasContext *ctx, int indx) { return 24 + (indx % 4) * 2; }
>> >> +static int to_D(DisasContext *ctx, int indx) { return (indx % 16) * 2; }
>> >> +
>> >> +static uint16_t next_word(DisasContext *ctx)
>> >> +{
>> >> +    return cpu_lduw_code(ctx->env, ctx->npc++ * 2);
>> >> +}
>> >> +
>> >> +static int append_16(DisasContext *ctx, int x)
>> >> +{
>> >> +    return x << 16 | next_word(ctx);
>> >> +}
>> >> +
>> >> +
>> >> +static bool avr_have_feature(DisasContext *ctx, int feature)
>> >> +{
>> >> +    if (!avr_feature(ctx->env, feature)) {
>> >> +        gen_helper_unsupported(cpu_env);
>> >> +        ctx->bstate = DISAS_NORETURN;
>> >> +        return false;
>> >> +    }
>> >> +    return true;
>> >> +}
>> >> +
>> >> +static bool decode_insn(DisasContext *ctx, uint16_t insn);
>> >> +#include "decode_insn.inc.c"
>> >> +
>> >> --
>> >> 2.17.2 (Apple Git-113)
>> >>
>>
>>
>> --
>> Best Regards,
>> Michael Rolnik



-- 
Best Regards,
Michael Rolnik


  reply	other threads:[~2019-10-17 20:44 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-14 16:18 [PATCH v32 00/13] QEMU AVR 8 bit cores Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 01/13] target/avr: Add outward facing interfaces and core CPU logic Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 02/13] target/avr: Add instruction helpers Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 03/13] target/avr: Add instruction decoding Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 04/13] target/avr: Add instruction translation - Registers definition Michael Rolnik
2019-10-17 17:25   ` Aleksandar Markovic
2019-10-17 19:15     ` Michael Rolnik
2019-10-17 20:16       ` Aleksandar Markovic
2019-10-17 20:43         ` Michael Rolnik [this message]
2019-10-18  8:52           ` Aleksandar Markovic
2019-10-18 11:27             ` Michael Rolnik
2019-10-18 12:10               ` Michael Rolnik
2019-10-18 13:23               ` Aleksandar Markovic
2019-10-18 13:56                 ` Michael Rolnik
2019-10-18 16:51                   ` Aleksandar Markovic
2019-10-18 18:08                     ` Aleksandar Markovic
2019-10-19 20:18                       ` Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 05/13] target/avr: Add instruction translation - Arithmetic and Logic Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 06/13] target/avr: Add instruction translation - Branch Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 07/13] target/avr: Add instruction translation - Bit and Bit-test Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 08/13] target/avr: Add instruction translation - MCU Control Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 09/13] target/avr: Add instruction translation - CPU main translation function Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 10/13] target/avr: Add limited support for USART and 16 bit timer peripherals Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 11/13] target/avr: Add example board configuration Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 12/13] target/avr: Register AVR support with the rest of QEMU, the build system, and the WMAINTAINERS file Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 13/13] target/avr: Add tests Michael Rolnik
  -- strict thread matches above, loose matches on Subject: below --
2019-10-13  7:47 [PATCH v32 00/13] QEMU AVR 8 bit cores Michael Rolnik
2019-10-13  7:47 ` [PATCH v32 04/13] target/avr: Add instruction translation - Registers definition Michael Rolnik

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=CAK4993jaqFrzcmAyz7neJoq5cQgnB_o35dZhdh9_uVyOfKGG-Q@mail.gmail.com \
    --to=mrolnik@gmail.com \
    --cc=aleksandar.m.mail@gmail.com \
    --cc=dovgaluk@ispras.ru \
    --cc=imammedo@redhat.com \
    --cc=philmd@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=thuth@redhat.com \
    /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).