From: joeyli <jlee@suse.com>
To: Matt Fleming <matt@codeblueprint.co.uk>
Cc: "Lee, Chun-Yi" <joeyli.kernel@gmail.com>,
linux-kernel@vger.kernel.org, linux-efi@vger.kernel.org,
linux-pm@vger.kernel.org, "Rafael J. Wysocki" <rjw@rjwysocki.net>,
Matthew Garrett <matthew.garrett@nebula.com>,
Len Brown <len.brown@intel.com>, Pavel Machek <pavel@ucw.cz>,
Josh Boyer <jwboyer@redhat.com>, Vojtech Pavlik <vojtech@suse.cz>,
Matt Fleming <matt.fleming@intel.com>,
Jiri Kosina <jkosina@suse.cz>, "H. Peter Anvin" <hpa@zytor.com>,
Ingo Molnar <mingo@redhat.com>
Subject: Re: [PATCH v2 08/16] x86/efi: Carrying hibernation key by setup data
Date: Thu, 27 Aug 2015 17:28:26 +0800 [thread overview]
Message-ID: <20150827092826.GD27415@linux-rxt1.site> (raw)
In-Reply-To: <20150821124026.GB3310@codeblueprint.co.uk>
On Fri, Aug 21, 2015 at 01:40:26PM +0100, Matt Fleming wrote:
> On Tue, 11 Aug, at 02:16:28PM, Lee, Chun-Yi wrote:
> > For forwarding hibernation key from EFI stub to boot kernel, this patch
> > allocates setup data for carrying hibernation key, size and the status
> > of efi operating.
>
> This could do with some more information, and include that the key is
> used to validate hibernate images.
>
> But now that I think about it, is there a reason this patch hasn't
> been merged with patch 6? The memory leak I mentioned in patch 6
> becomes a non-issue in this one, so it would be good if these two
> could be squashed together.
>
OK, I will merge this patch with patch 6.
Actually the sequence of patches are from the order of my developing.
And, the purpose of code in this patch a bit different with patch 6,
so I didn't merge them together.
> > Reviewed-by: Jiri Kosina <jkosina@suse.com>
> > Tested-by: Jiri Kosina <jkosina@suse.com>
> > Signed-off-by: Lee, Chun-Yi <jlee@suse.com>
> > ---
> > arch/x86/boot/compressed/eboot.c | 26 +++++++++++++++++++++++---
> > arch/x86/include/uapi/asm/bootparam.h | 1 +
> > 2 files changed, 24 insertions(+), 3 deletions(-)
> >
> > diff --git a/arch/x86/boot/compressed/eboot.c b/arch/x86/boot/compressed/eboot.c
> > index 463aa9b..c838d09 100644
> > --- a/arch/x86/boot/compressed/eboot.c
> > +++ b/arch/x86/boot/compressed/eboot.c
> > @@ -1394,18 +1394,22 @@ static void setup_hibernation_keys(struct boot_params *params)
> > {
> > unsigned long key_size;
> > unsigned long attributes;
> > + struct setup_data *setup_data, *hibernation_setup_data;
> > struct hibernation_keys *keys;
> > + unsigned long size = 0;
> > efi_status_t status;
>
> One thing to be aware of is that eboot.c has mainly used the
> "reverse-christmas-tree" style of variable declarations, with longer
> lines first, and shorter ones following. I haven't mentioned it before
> because none of your changes seemed to be too different (and it's not
> a tree-wide convention), but the above setup_data line goes a bit too
> far.
>
> Could you try and keep them ordered, longest line first?
>
Sure, sorry for I didn't aware that before.
> >
> > /* Allocate setup_data to carry keys */
> > + size = sizeof(struct setup_data) + sizeof(struct hibernation_keys);
> > status = efi_call_early(allocate_pool, EFI_LOADER_DATA,
> > - sizeof(struct hibernation_keys), &keys);
> > + size, &hibernation_setup_data);
> > if (status != EFI_SUCCESS) {
> > efi_printk(sys_table, "Failed to alloc mem for hibernation keys\n");
> > return;
> > }
> >
> > - memset(keys, 0, sizeof(struct hibernation_keys));
> > + memset(hibernation_setup_data, 0, size);
> > + keys = (struct hibernation_keys *) hibernation_setup_data->data;
> >
> > status = efi_call_early(get_variable, HIBERNATION_KEY,
> > &EFI_HIBERNATION_GUID, &attributes,
> > @@ -1419,7 +1423,8 @@ static void setup_hibernation_keys(struct boot_params *params)
> > if (status == EFI_SUCCESS) {
> > efi_printk(sys_table, "Cleaned existing hibernation key\n");
> > status = EFI_NOT_FOUND;
> > - }
> > + } else
> > + goto clean_fail;
>
> Please add braces for the 'else' clause. Also, please include a
> comment stating that the reason you jump to the label instead of
> returning is so that the EFI status error code can be recorded in
> hibernation_setup_data.
>
Thanks for suggestions, I will modify it.
> > }
> >
> > if (status != EFI_SUCCESS) {
> > @@ -1436,6 +1441,21 @@ static void setup_hibernation_keys(struct boot_params *params)
> > if (status != EFI_SUCCESS)
> > efi_printk(sys_table, "Failed to set hibernation key\n");
> > }
> > +
> > +clean_fail:
> > + hibernation_setup_data->type = SETUP_HIBERNATION_KEYS;
> > + hibernation_setup_data->len = sizeof(struct hibernation_keys);
> > + hibernation_setup_data->next = 0;
> > + keys->hkey_status = efi_status_to_err(status);
> > +
> > + setup_data = (struct setup_data *)params->hdr.setup_data;
> > + while (setup_data && setup_data->next)
> > + setup_data = (struct setup_data *)setup_data->next;
> > +
> > + if (setup_data)
> > + setup_data->next = (unsigned long)hibernation_setup_data;
> > + else
> > + params->hdr.setup_data = (unsigned long)hibernation_setup_data;
>
> This label name is a little confusing because you reach it both when
> the EFI boot variable was successfully created and when a bogus EFI
> variable failed to be deleted, i.e. it's not always reached because of
> a failure.
>
> How about 'setup' or simply 'out' ?
>
I will change the label to 'setup' that match with setting setup_data.
> --
> Matt Fleming, Intel Open Source Technology Center
Thanks a lot!
Joey Lee
next prev parent reply other threads:[~2015-08-27 9:28 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-08-11 6:16 [PATCH v2 00/16] Signature verification of hibernate snapshot Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 01/16] PM / hibernate: define HMAC algorithm and digest size of hibernation Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 02/16] x86/efi: Add get and set variable to EFI services pointer table Lee, Chun-Yi
2015-08-19 16:35 ` Matt Fleming
2015-08-11 6:16 ` [PATCH v2 03/16] x86/boot: Public getting random boot function Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 04/16] x86/efi: Generating random number in EFI stub Lee, Chun-Yi
2015-08-20 14:12 ` Matt Fleming
2015-08-27 4:06 ` joeyli
2015-08-11 6:16 ` [PATCH v2 05/16] x86/efi: Get entropy through EFI random number generator protocol Lee, Chun-Yi
2015-08-20 14:47 ` Matt Fleming
2015-08-27 4:51 ` joeyli
2015-08-20 20:26 ` Matt Fleming
2015-08-27 6:17 ` joeyli
2015-08-11 6:16 ` [PATCH v2 06/16] x86/efi: Generating random HMAC key for siging hibernate image Lee, Chun-Yi
2015-08-20 20:40 ` Matt Fleming
2015-08-27 9:04 ` joeyli
2015-09-09 12:15 ` Matt Fleming
2015-09-13 2:47 ` joeyli
2015-08-11 6:16 ` [PATCH v2 07/16] efi: Make efi_status_to_err() public Lee, Chun-Yi
2015-08-20 15:07 ` Matt Fleming
2015-08-27 9:06 ` joeyli
2015-08-11 6:16 ` [PATCH v2 08/16] x86/efi: Carrying hibernation key by setup data Lee, Chun-Yi
2015-08-15 17:07 ` Pavel Machek
2015-08-16 5:28 ` joeyli
2015-08-16 21:23 ` Jiri Kosina
2015-08-17 6:54 ` Nigel Cunningham
2015-08-21 12:40 ` Matt Fleming
2015-08-27 9:28 ` joeyli [this message]
2015-08-11 6:16 ` [PATCH v2 09/16] PM / hibernate: Reserve hibernation key and erase footprints Lee, Chun-Yi
2015-08-13 2:45 ` Chen, Yu C
2015-08-13 3:25 ` joeyli
2015-08-13 14:33 ` joeyli
2015-08-21 13:27 ` Matt Fleming
2015-08-27 10:21 ` joeyli
2015-09-09 12:24 ` Matt Fleming
2015-09-13 2:58 ` joeyli
2015-08-11 6:16 ` [PATCH v2 10/16] PM / hibernate: Generate and verify signature of hibernate snapshot Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 11/16] PM / hibernate: Avoid including hibernation key to hibernate image Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 12/16] PM / hibernate: Forward signature verifying result and key to image kernel Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 13/16] PM / hibernate: Add configuration to enforce signature verification Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 14/16] PM / hibernate: Allow user trigger hibernation key re-generating Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 15/16] PM / hibernate: Bypass verification logic on legacy BIOS Lee, Chun-Yi
2015-08-11 6:16 ` [PATCH v2 16/16] PM / hibernate: Document signature verification of hibernate snapshot Lee, Chun-Yi
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=20150827092826.GD27415@linux-rxt1.site \
--to=jlee@suse.com \
--cc=hpa@zytor.com \
--cc=jkosina@suse.cz \
--cc=joeyli.kernel@gmail.com \
--cc=jwboyer@redhat.com \
--cc=len.brown@intel.com \
--cc=linux-efi@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=matt.fleming@intel.com \
--cc=matt@codeblueprint.co.uk \
--cc=matthew.garrett@nebula.com \
--cc=mingo@redhat.com \
--cc=pavel@ucw.cz \
--cc=rjw@rjwysocki.net \
--cc=vojtech@suse.cz \
/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).