From: Stanislav Kholmanskikh <stanislav.kholmanskikh@oracle.com>
To: The development of GNU GRUB <grub-devel@gnu.org>
Cc: vasily.isaenko@oracle.com
Subject: Re: [PATCH 1/2] ieee1275: alloc-mem and free-mem
Date: Fri, 18 Nov 2016 15:32:43 +0300 [thread overview]
Message-ID: <582EF4EB.1040305@oracle.com> (raw)
In-Reply-To: <20161115212243.GH16470@router-fw-old.local.net-space.pl>
Hi, Daniel.
Thank you for review.
My comments are below.
On 11/16/2016 12:22 AM, Daniel Kiper wrote:
> On Tue, Apr 12, 2016 at 03:39:55PM +0300, Stanislav Kholmanskikh wrote:
>> Add wrappers for memory allocation using
>> alloc-mem and free-mem commands from the User Interface.
>
> Please tell why it is needed. Additionally, please forgive me if it is stupid
> question, why are you using command line to allocate/free memory? There is
> a lack of better API in IEEE 1275?
In the current git code in grub-core/net/drivers/ieee1275/ofnet.c the
search_net_devices() function uses the "alloc-mem" command for
allocation of the transmit buffer for the case when
GRUB_IEEE1275_FLAG_VIRT_TO_REAL_BROKEN is set.
Yes, besides "alloc-mem", "free-mem", there are other options for
allocating memory. But, to be honest, I don't know why grub_malloc()
cannot be used for the receive buffer on those systems. From the flag's
name it seems those systems don't have a MMU. Ok, but grub_malloc() is
called from many places in grub, and, presumably, grub works on those
systems. I don't have such a MacRISC system to try grub_malloc() for the
buffers (grub-core/kern/ieee1275/cmain.c).
In patch 2 I'm implementing a receive buffer, so I decided to keep this
case as is, but move "alloc-mem", "free-mem" into functions.
Does it answer your questions? If yes, I'll update the commit message
for V2 with a more detailed explanation why these functions are needed.
>
>> Signed-off-by: Stanislav Kholmanskikh <stanislav.kholmanskikh@oracle.com>
>> ---
>> grub-core/kern/ieee1275/openfw.c | 68 ++++++++++++++++++++++++++++++++++++++
>> include/grub/ieee1275/ieee1275.h | 2 +
>> 2 files changed, 70 insertions(+), 0 deletions(-)
>>
>> diff --git a/grub-core/kern/ieee1275/openfw.c b/grub-core/kern/ieee1275/openfw.c
>> index ddb7783..35225ec 100644
>> --- a/grub-core/kern/ieee1275/openfw.c
>> +++ b/grub-core/kern/ieee1275/openfw.c
>> @@ -561,3 +561,71 @@ grub_ieee1275_canonicalise_devname (const char *path)
>> return NULL;
>> }
>>
>> +/* Allocate memory with alloc-mem */
>> +void *
>> +grub_ieee1275_alloc_mem (grub_size_t len)
>> +{
>> + struct alloc_args
>> + {
>> + struct grub_ieee1275_common_hdr common;
>> + grub_ieee1275_cell_t method;
>> + grub_ieee1275_cell_t len;
>> + grub_ieee1275_cell_t catch;
>> + grub_ieee1275_cell_t result;
>> + }
>> + args;
>> +
>> + if (grub_ieee1275_test_flag (GRUB_IEEE1275_FLAG_CANNOT_INTERPRET))
>> + {
>> + grub_error (GRUB_ERR_UNKNOWN_COMMAND, N_("interpret is not supported"));
>> + return NULL;
>> + }
>> +
>> + INIT_IEEE1275_COMMON (&args.common, "interpret", 2, 2);
>> + args.len = len;
>> + args.method = (grub_ieee1275_cell_t) "alloc-mem";
>> +
>> + if (IEEE1275_CALL_ENTRY_FN (&args) == -1
>> + || args.catch)
>
> I think that this can be in one line.
Ok.
>
>> + {
>> + grub_error (GRUB_ERR_INVALID_COMMAND, N_("alloc-mem failed"));
>> + return NULL;
>> + }
>> + else
>> + return (void *)args.result;
>> +}
>> +
>> +/* Free memory allocated by alloc-mem */
>> +grub_err_t
>> +grub_ieee1275_free_mem (void *addr, grub_size_t len)
>> +{
>> + struct free_args
>> + {
>> + struct grub_ieee1275_common_hdr common;
>> + grub_ieee1275_cell_t method;
>> + grub_ieee1275_cell_t len;
>> + grub_ieee1275_cell_t addr;
>> + grub_ieee1275_cell_t catch;
>> + }
>> + args;
>> +
>> + if (grub_ieee1275_test_flag (GRUB_IEEE1275_FLAG_CANNOT_INTERPRET))
>> + {
>> + grub_error (GRUB_ERR_UNKNOWN_COMMAND, N_("interpret is not supported"));
>> + return grub_errno;
>> + }
>> +
>> + INIT_IEEE1275_COMMON (&args.common, "interpret", 3, 1);
>> + args.addr = (grub_ieee1275_cell_t)addr;
>> + args.len = len;
>> + args.method = (grub_ieee1275_cell_t) "free-mem";
>> +
>> + if (IEEE1275_CALL_ENTRY_FN(&args) == -1
>> + || args.catch)
>
> Ditto.
Ok.
>
> Daniel
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel
>
next prev parent reply other threads:[~2016-11-18 12:33 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-04-12 12:39 [PATCH 1/2] ieee1275: alloc-mem and free-mem Stanislav Kholmanskikh
2016-04-12 12:39 ` [PATCH 2/2] ofnet: implement a receive buffer Stanislav Kholmanskikh
2016-05-10 10:45 ` Stanislav Kholmanskikh
2016-07-13 14:35 ` Stanislav Kholmanskikh
2016-10-13 8:13 ` Stanislav Kholmanskikh
2016-11-15 22:34 ` Daniel Kiper
2016-11-18 13:29 ` Stanislav Kholmanskikh
2016-11-21 21:48 ` Daniel Kiper
2016-11-22 14:08 ` Stanislav Kholmanskikh
2016-11-23 11:16 ` Daniel Kiper
2016-11-23 15:09 ` Stanislav Kholmanskikh
2016-11-24 9:25 ` Daniel Kiper
2016-11-30 15:27 ` Stanislav Kholmanskikh
2016-11-15 21:22 ` [PATCH 1/2] ieee1275: alloc-mem and free-mem Daniel Kiper
2016-11-18 12:32 ` Stanislav Kholmanskikh [this message]
2017-05-11 1:51 ` Vladimir 'phcoder' Serbinenko
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=582EF4EB.1040305@oracle.com \
--to=stanislav.kholmanskikh@oracle.com \
--cc=grub-devel@gnu.org \
--cc=vasily.isaenko@oracle.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).