From: Sergei Shtylyov <sshtylyov@ru.mvista.com>
To: petkovbb@gmail.com, Sergei Shtylyov <sshtylyov@ru.mvista.com>,
Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>,
linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 5/6] ide: remove ide_execute_pkt_cmd()
Date: Mon, 16 Feb 2009 02:18:01 +0300 [thread overview]
Message-ID: <4998A2A9.2060607@ru.mvista.com> (raw)
In-Reply-To: <20090215173901.GA5156@gollum.tnic>
Hello.
Borislav Petkov wrote:
>> I still don't undestand why you assume that such variable will be
>> alloceted on stack -- gcc has 3 registers available for local variables
>> (which doesn't have to save across function calls). However, the
>> register variables have to take stack space indeed as they need to be
>> saved on funciton entry... though I'm not sure that gcc will necessary
>> put such variable in one of those 3 registers if it figures out that
>> there are no function calls going to happen during its life time.
>>
>>> and the code is more readable. A win-win situation, I'd say :).
>>>
>> You haven't presented the code which gets generated when the local
>> variable is used, so it's impossible to compare.
>>
>
> Here's another example from ide-disk.c where you have stack variables cashing
> those flags checks:
>
They are caching the result of !! in 'u8' variables -- which is not
the same as cahing the flags. I suspect gcc avoid putting byte-sized
variables into registers...
> <ide-disk.c>
> static ide_startstop_t __ide_do_rw_disk(ide_drive_t *drive, struct request *rq,
> sector_t block)
> {
> ide_hwif_t *hwif = drive->hwif;
> u16 nsectors = (u16)rq->nr_sectors;
> u8 lba48 = !!(drive->dev_flags & IDE_DFLAG_LBA48);
> u8 dma = !!(drive->dev_flags & IDE_DFLAG_USING_DMA);
> ide_task_t task;
> struct ide_taskfile *tf = &task.tf;
> ide_startstop_t rc;
>
> if ((hwif->host_flags & IDE_HFLAG_NO_LBA48_DMA) && lba48 && dma) {
> if (block + rq->nr_sectors > 1ULL << 28)
> dma = 0;
> else
> lba48 = 0;
> }
> </ide-disk.c>
>
> Corresponding asm (this time i386 but I don't think it matters since we
> need at least one arch to prove my point).
>
> <ide-disk.s>
> .LVL48:
> .loc 1 94 0
> movl %eax, %edx # D.32119, tmp93
> .loc 1 95 0
> shrl %eax # D.32119
>
Where's the shift count I wonder?
> andb $1, %al #,
> movb %al, -58(%ebp) #, dma
> .LVL49:
> .loc 1 100 0
> movl -52(%ebp), %eax # hwif,
> .loc 1 94 0
> shrl $21, %edx #, tmp93
> andb $1, %dl #,
> movb %dl, -57(%ebp) #, lba48
> .LVL50:
> .loc 1 100 0
> testb $4, 90(%eax) #, <variable>.host_flags
> je .L37 #,
> testb %dl, %dl #
> je .L37 #,
> cmpb $0, -58(%ebp) # dma
> movb $1, -57(%ebp) #, lba48
> .LVL51:
> </ide-disk.s>
>
> Now look at the last lines at labels .LVL48 and .LVL49 - they both save
> those 1-byte u8's called dma and lba48 on the stack at -57(%ebp) and
> -58(%ebp), respectively. And guess what, later on label LVL50 they get
> accessed in the check.
If you look better, you'll see that the copy of 'lba48' in the %dl
register gets used.
> And several times more later, which in most sane
> architectures still means cache accesses but when you have registers its
> even faster :).
Didn't quite get that statement.
Well, this example wasn't very convincing...
MBR. Sergei
next prev parent reply other threads:[~2009-02-15 23:18 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-02-09 23:19 [PATCH 0/6] ide: more unifications of ATA and ATAPI support Bartlomiej Zolnierkiewicz
2009-02-09 23:19 ` [PATCH 1/6] ide: pass command to ide_map_sg() Bartlomiej Zolnierkiewicz
2009-02-11 6:36 ` Borislav Petkov
2009-02-11 16:28 ` Bartlomiej Zolnierkiewicz
2009-02-09 23:19 ` [PATCH 2/6] ide: use do_rw_taskfile() for ATA_CMD_PACKET commands Bartlomiej Zolnierkiewicz
2009-02-09 23:20 ` [PATCH 3/6] ide: set hwif->expiry prior to calling [__]ide_set_handler() Bartlomiej Zolnierkiewicz
2009-03-16 14:23 ` Sergei Shtylyov
2009-02-09 23:20 ` [PATCH 4/6] ide: add ->dma_expiry method and remove ->dma_exec_cmd one Bartlomiej Zolnierkiewicz
2009-02-10 18:18 ` Sergei Shtylyov
2009-02-11 16:30 ` Bartlomiej Zolnierkiewicz
2009-02-11 17:30 ` Sergei Shtylyov
2009-02-17 14:16 ` Bartlomiej Zolnierkiewicz
2009-02-17 14:29 ` Sergei Shtylyov
2009-02-09 23:20 ` [PATCH 5/6] ide: remove ide_execute_pkt_cmd() Bartlomiej Zolnierkiewicz
2009-02-11 6:55 ` Borislav Petkov
2009-02-11 13:22 ` Sergei Shtylyov
2009-02-11 13:37 ` Borislav Petkov
2009-02-11 13:49 ` Sergei Shtylyov
2009-02-11 16:32 ` Borislav Petkov
2009-02-15 12:24 ` Sergei Shtylyov
2009-02-15 17:39 ` Borislav Petkov
2009-02-15 23:18 ` Sergei Shtylyov [this message]
2009-02-16 8:56 ` Borislav Petkov
2009-02-11 16:37 ` Bartlomiej Zolnierkiewicz
2009-02-09 23:20 ` [PATCH 6/6] ide: keep track of number of bytes instead of sectors in struct ide_cmd Bartlomiej Zolnierkiewicz
2009-02-11 7:16 ` [PATCH 0/6] ide: more unifications of ATA and ATAPI support Borislav Petkov
2009-02-23 22:51 ` Bartlomiej Zolnierkiewicz
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=4998A2A9.2060607@ru.mvista.com \
--to=sshtylyov@ru.mvista.com \
--cc=bzolnier@gmail.com \
--cc=linux-ide@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=petkovbb@gmail.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.