From: Dave Hansen <dave.hansen@intel.com>
To: "Kirill A. Shutemov" <kirill@shutemov.name>,
Borislav Petkov <bp@alien8.de>, Andy Lutomirski <luto@kernel.org>,
Sean Christopherson <seanjc@google.com>,
Andrew Morton <akpm@linux-foundation.org>,
Joerg Roedel <jroedel@suse.de>
Cc: Andi Kleen <ak@linux.intel.com>,
Kuppuswamy Sathyanarayanan
<sathyanarayanan.kuppuswamy@linux.intel.com>,
David Rientjes <rientjes@google.com>,
Vlastimil Babka <vbabka@suse.cz>,
Tom Lendacky <thomas.lendacky@amd.com>,
Thomas Gleixner <tglx@linutronix.de>,
Peter Zijlstra <peterz@infradead.org>,
Paolo Bonzini <pbonzini@redhat.com>,
Ingo Molnar <mingo@redhat.com>,
Varad Gautam <varad.gautam@suse.com>,
Dario Faggioli <dfaggioli@suse.com>,
x86@kernel.org, linux-mm@kvack.org, linux-coco@lists.linux.dev,
linux-kernel@vger.kernel.org,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Subject: Re: [PATCH 2/5] efi/x86: Implement support for unaccepted memory
Date: Tue, 10 Aug 2021 10:50:33 -0700 [thread overview]
Message-ID: <142bccc6-0e67-dfc1-9069-b773c2bad585@intel.com> (raw)
In-Reply-To: <20210810062626.1012-3-kirill.shutemov@linux.intel.com>
...
> +void mark_unaccepted(struct boot_params *params, u64 start, u64 num)
> +{
Some of these interfaces like accept_memory() take a start/end physical
address. Having this take a "num pages" is bound to cause confusion.
Could you make these all consistently take start/end physical addresses?
> + u64 end = start + num * PAGE_SIZE;
> + unsigned int npages;
Could you comment those, please?
/*
* The accepted memory bitmap only works at PMD_SIZE
* granularity. If a request comes in to mark memory
* as unaccepted which is not PMD_SIZE-aligned, simply
* accept the memory now since it can not be *marked* as
* unaccepted.
*/
Then go on to comment the three cases:
/* Check for ranges which do not span a whole PMD_SIZE area: */
> + if ((start & PMD_MASK) == (end & PMD_MASK)) {
> + npages = (end - start) / PAGE_SIZE;
> + __accept_memory(start, start + npages * PAGE_SIZE);
> + return;
> + }
Hmm, is it possible to have this case hit, but neither of the two below
cases? This seems to be looking for a case where the range is somehow
entirely contained in one PMD_SIZE area, but where it doesn't consume a
whole area.
Wouldn't that mean that 'start' or 'end' must be unaligned?
> + if (start & ~PMD_MASK) {
> + npages = (round_up(start, PMD_SIZE) - start) / PAGE_SIZE;
> + __accept_memory(start, start + npages * PAGE_SIZE);
> + start = round_up(start, PMD_SIZE);
> + }
> +
> + if (end & ~PMD_MASK) {
> + npages = (end - round_down(end, PMD_SIZE)) / PAGE_SIZE;
> + end = round_down(end, PMD_SIZE);
> + __accept_memory(end, end + npages * PAGE_SIZE);
> + }
> + npages = (end - start) / PMD_SIZE;
> + bitmap_set((unsigned long *)params->unaccepted_memory,
> + start / PMD_SIZE, npages);
> +}
Even though it's changed right there, it's a bit cruel to change the
units of 'npages' right in the middle of a function. It's just asking
for bugs.
It would only take a single extra variable declaration to make this
unambiguous:
u64 nr_unaccepted_bits;
or something, then you can do:
nr_unaccepted_bits = (end - start) / PMD_SIZE;
bitmap_set((unsigned long *)params->unaccepted_memory,
start / PMD_SIZE, nr_unaccepted_bits);
...
> static efi_status_t allocate_e820(struct boot_params *params,
> + struct efi_boot_memmap *map,
> struct setup_data **e820ext,
> u32 *e820ext_size)
> {
> - unsigned long map_size, desc_size, map_key;
> efi_status_t status;
> - __u32 nr_desc, desc_version;
> -
> - /* Only need the size of the mem map and size of each mem descriptor */
> - map_size = 0;
> - status = efi_bs_call(get_memory_map, &map_size, NULL, &map_key,
> - &desc_size, &desc_version);
> - if (status != EFI_BUFFER_TOO_SMALL)
> - return (status != EFI_SUCCESS) ? status : EFI_UNSUPPORTED;
I noticed that there's no reference to EFI_BUFFER_TOO_SMALL in the hunks
you added back. That makes me a bit nervous that this is going to
unintentionally change behavior.
It might be worth having a preparatory reorganization patch for
allocate_e820() before this new feature is added to make this more clear.
> + __u32 nr_desc;
> + bool unaccepted_memory_present = false;
> + u64 max_addr = 0;
> + int i;
>
> - nr_desc = map_size / desc_size + EFI_MMAP_NR_SLACK_SLOTS;
> + status = efi_get_memory_map(map);
> + if (status != EFI_SUCCESS)
> + return status;
>
> - if (nr_desc > ARRAY_SIZE(params->e820_table)) {
> - u32 nr_e820ext = nr_desc - ARRAY_SIZE(params->e820_table);
> + nr_desc = *map->map_size / *map->desc_size;
> + if (nr_desc > ARRAY_SIZE(params->e820_table) - EFI_MMAP_NR_SLACK_SLOTS) {
> + u32 nr_e820ext = nr_desc - ARRAY_SIZE(params->e820_table) -
> + EFI_MMAP_NR_SLACK_SLOTS;
>
> status = alloc_e820ext(nr_e820ext, e820ext, e820ext_size);
> if (status != EFI_SUCCESS)
> return status;
> }
>
> + if (!IS_ENABLED(CONFIG_UNACCEPTED_MEMORY))
> + return EFI_SUCCESS;
> +
> + /* Check if there's any unaccepted memory and find the max address */
> + for (i = 0; i < nr_desc; i++) {
> + efi_memory_desc_t *d;
> +
> + d = efi_early_memdesc_ptr(*map->map, *map->desc_size, i);
> + if (d->type == EFI_UNACCEPTED_MEMORY)
> + unaccepted_memory_present = true;
> + if (d->phys_addr + d->num_pages * PAGE_SIZE > max_addr)
> + max_addr = d->phys_addr + d->num_pages * PAGE_SIZE;
> + }
This 'max_addr' variable looks a bit funky.
It *seems* like it's related only to EFI_UNACCEPTED_MEMORY, but it's not
underneath the EFI_UNACCEPTED_MEMORY check. Is this somehow assuming
that once unaccepted memory as been found that *all* memory found in
later descriptors at higher addresses is also going to be unaccepted?
> + /*
> + * If unaccepted memory present allocate a bitmap to track what memory
> + * has to be accepted before access.
> + *
> + * One bit in the bitmap represents 2MiB in the address space: one 4k
> + * page is enough to track 64GiB or physical address space.
> + *
> + * In the worst case scenario -- a huge hole in the middle of the
> + * address space -- we would need 256MiB to handle 4PiB of the address
> + * space.
> + *
> + * TODO: handle situation if params->unaccepted_memory has already set.
> + * It's required to deal with kexec.
> + */
> + if (unaccepted_memory_present) {
> + unsigned long *unaccepted_memory = NULL;
> + u64 size = DIV_ROUND_UP(max_addr, PMD_SIZE * BITS_PER_BYTE);
Oh, so the bitmap has to be present for *all* memory, not just
unaccepted memory. So, we really do need to know the 'max_addr' so that
we can allocate the bitmap for so that can be marked in the bitmap has
having been accepted.
> + status = efi_allocate_pages(size,
> + (unsigned long *)&unaccepted_memory,
> + ULONG_MAX);
> + if (status != EFI_SUCCESS)
> + return status;
> + memset(unaccepted_memory, 0, size);
> + params->unaccepted_memory = (u64)unaccepted_memory;
> + }
It might be nice to refer to setup_e820() here to mention that it is the
thing that actually fills out the bitmap.
next prev parent reply other threads:[~2021-08-10 17:50 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-10 6:26 [PATCH 0/5] x86: Impplement support for unaccepted memory Kirill A. Shutemov
2021-08-10 6:26 ` [PATCH 1/5] mm: Add " Kirill A. Shutemov
2021-08-10 7:48 ` David Hildenbrand
2021-08-10 15:02 ` Kirill A. Shutemov
2021-08-10 15:21 ` David Hildenbrand
2021-08-12 20:34 ` Kirill A. Shutemov
2021-08-10 18:13 ` Dave Hansen
2021-08-10 18:30 ` Andi Kleen
2021-08-10 18:56 ` Dave Hansen
2021-08-10 19:23 ` Andi Kleen
2021-08-10 19:46 ` Dave Hansen
2021-08-10 21:20 ` Andi Kleen
2021-08-12 8:19 ` Joerg Roedel
2021-08-12 14:14 ` Dave Hansen
2021-08-12 20:49 ` Kirill A. Shutemov
2021-08-12 20:59 ` Dave Hansen
2021-08-12 21:23 ` Kirill A. Shutemov
2021-08-13 14:49 ` Joerg Roedel
2021-08-17 15:00 ` David Hildenbrand
2021-08-19 9:55 ` Joerg Roedel
2021-08-19 10:06 ` David Hildenbrand
2021-08-10 20:50 ` Dave Hansen
2021-08-12 21:08 ` Kirill A. Shutemov
2021-08-10 6:26 ` [PATCH 2/5] efi/x86: Implement " Kirill A. Shutemov
2021-08-10 17:50 ` Dave Hansen [this message]
2021-08-12 21:14 ` Kirill A. Shutemov
2021-08-12 21:43 ` Dave Hansen
2021-08-10 18:30 ` Dave Hansen
2021-08-10 19:08 ` Kirill A. Shutemov
2021-08-10 19:19 ` Dave Hansen
2021-08-12 21:17 ` Kirill A. Shutemov
2021-08-10 6:26 ` [PATCH 3/5] x86/boot/compressed: Handle " Kirill A. Shutemov
2021-08-10 6:26 ` [PATCH 4/5] x86/mm: Provide helpers for " Kirill A. Shutemov
2021-08-10 18:16 ` Dave Hansen
2021-08-12 20:31 ` Kirill A. Shutemov
2021-08-10 6:26 ` [PATCH 5/5] x86/tdx: Unaccepted memory support Kirill A. Shutemov
2021-08-10 14:08 ` [PATCH 0/5] x86: Impplement support for unaccepted memory Dave Hansen
2021-08-10 15:15 ` Kirill A. Shutemov
2021-08-10 15:51 ` Dave Hansen
2021-08-10 17:31 ` Kirill A. Shutemov
2021-08-10 17:36 ` Dave Hansen
2021-08-10 17:51 ` Kirill A. Shutemov
2021-08-10 18:19 ` Dave Hansen
2021-08-10 18:39 ` Kirill A. Shutemov
2021-08-12 8:23 ` Joerg Roedel
2021-08-12 10:10 ` Kirill A. Shutemov
2021-08-12 19:33 ` Andi Kleen
2021-08-12 20:22 ` Kirill A. Shutemov
2021-08-13 14:56 ` Joerg Roedel
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=142bccc6-0e67-dfc1-9069-b773c2bad585@intel.com \
--to=dave.hansen@intel.com \
--cc=ak@linux.intel.com \
--cc=akpm@linux-foundation.org \
--cc=bp@alien8.de \
--cc=dfaggioli@suse.com \
--cc=jroedel@suse.de \
--cc=kirill.shutemov@linux.intel.com \
--cc=kirill@shutemov.name \
--cc=linux-coco@lists.linux.dev \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=luto@kernel.org \
--cc=mingo@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=rientjes@google.com \
--cc=sathyanarayanan.kuppuswamy@linux.intel.com \
--cc=seanjc@google.com \
--cc=tglx@linutronix.de \
--cc=thomas.lendacky@amd.com \
--cc=varad.gautam@suse.com \
--cc=vbabka@suse.cz \
--cc=x86@kernel.org \
/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).