[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v2 18/42] esp: accumulate SCSI commands for PDMA transfers in
From: |
Laurent Vivier |
Subject: |
Re: [PATCH v2 18/42] esp: accumulate SCSI commands for PDMA transfers in cmdbuf instead of pdma_buf |
Date: |
Tue, 2 Mar 2021 18:02:07 +0100 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.7.0 |
Le 09/02/2021 à 20:29, Mark Cave-Ayland a écrit :
> ESP SCSI commands are already accumulated in cmdbuf and so there is no need to
> keep a separate pdma_buf buffer. Accumulate SCSI commands for PDMA transfers
> in
> cmdbuf instead of pdma_buf so update cmdlen accordingly and change pdma_origin
> for PDMA transfers to CMD which allows the PDMA origin to be removed.
>
> This commit also removes a stray memcpy() from get_cmd() which is a no-op
> because
> cmdlen is always zero at the start of a command.
>
> Notionally the removal of pdma_buf from vmstate_esp_pdma also breaks migration
> compatibility for the PDMA subsection until its complete removal by the end of
> the series.
>
> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
> ---
> hw/scsi/esp.c | 56 +++++++++++++++++++------------------------
> include/hw/scsi/esp.h | 2 --
> 2 files changed, 25 insertions(+), 33 deletions(-)
>
> diff --git a/hw/scsi/esp.c b/hw/scsi/esp.c
> index 7134c0aff4..b846f022fb 100644
> --- a/hw/scsi/esp.c
> +++ b/hw/scsi/esp.c
> @@ -139,8 +139,6 @@ static void set_pdma(ESPState *s, enum pdma_origin_id
> origin,
> static uint8_t *get_pdma_buf(ESPState *s)
> {
> switch (s->pdma_origin) {
> - case PDMA:
> - return s->pdma_buf;
> case TI:
> return s->ti_buf;
> case CMD:
> @@ -161,14 +159,12 @@ static uint8_t esp_pdma_read(ESPState *s)
> }
>
> switch (s->pdma_origin) {
> - case PDMA:
> - val = s->pdma_buf[s->pdma_cur++];
> - break;
> case TI:
> val = s->ti_buf[s->pdma_cur++];
> break;
> case CMD:
> - val = s->cmdbuf[s->pdma_cur++];
> + val = s->cmdbuf[s->cmdlen++];
> + s->pdma_cur++;
> break;
> case ASYNC:
> val = s->async_buf[s->pdma_cur++];
> @@ -193,14 +189,12 @@ static void esp_pdma_write(ESPState *s, uint8_t val)
> }
>
> switch (s->pdma_origin) {
> - case PDMA:
> - s->pdma_buf[s->pdma_cur++] = val;
> - break;
> case TI:
> s->ti_buf[s->pdma_cur++] = val;
> break;
> case CMD:
> - s->cmdbuf[s->pdma_cur++] = val;
> + s->cmdbuf[s->cmdlen++] = val;
> + s->pdma_cur++;
> break;
> case ASYNC:
> s->async_buf[s->pdma_cur++] = val;
> @@ -256,8 +250,7 @@ static uint32_t get_cmd(ESPState *s, uint8_t *buf,
> uint8_t buflen)
> if (s->dma_memory_read) {
> s->dma_memory_read(s->dma_opaque, buf, dmalen);
> } else {
> - memcpy(s->pdma_buf, buf, dmalen);
> - set_pdma(s, PDMA, 0, dmalen);
> + set_pdma(s, CMD, 0, dmalen);
> esp_raise_drq(s);
> return 0;
> }
> @@ -316,24 +309,24 @@ static void satn_pdma_cb(ESPState *s)
> if (get_cmd_cb(s) < 0) {
> return;
> }
> - if (s->pdma_cur != s->pdma_start) {
> - do_cmd(s, get_pdma_buf(s) + s->pdma_start);
> + s->do_cmd = 0;
> + if (s->cmdlen) {
> + do_cmd(s, s->cmdbuf);
I don't understant how you can remove the pdma_start: normally it is here to
keep track of the
position in the pDMA if the migration is occuraing while a pDMA transfer.
Thanks,
Laurent
- Re: [PATCH v2 18/42] esp: accumulate SCSI commands for PDMA transfers in cmdbuf instead of pdma_buf,
Laurent Vivier <=