public inbox for kexec@lists.infradead.org
 help / color / mirror / Atom feed
From: Michael Ellerman <mpe@ellerman.id.au>
To: Hari Bathini <hbathini@linux.ibm.com>,
	Andrew Morton <akpm@linux-foundation.org>
Cc: Pingfan Liu <piliu@redhat.com>,
	Kexec-ml <kexec@lists.infradead.org>,
	Mimi Zohar <zohar@linux.ibm.com>,
	Nayna Jain <nayna@linux.ibm.com>, Petr Tesarik <ptesarik@suse.cz>,
	Mahesh J Salgaonkar <mahesh@linux.ibm.com>,
	Sourabh Jain <sourabhjain@linux.ibm.com>,
	lkml <linux-kernel@vger.kernel.org>,
	linuxppc-dev <linuxppc-dev@ozlabs.org>,
	Eric Biederman <ebiederm@xmission.com>,
	Thiago Jung Bauermann <bauerman@linux.ibm.com>,
	Dave Young <dyoung@redhat.com>, Vivek Goyal <vgoyal@redhat.com>
Subject: Re: [RESEND PATCH v5 06/11] ppc64/kexec_file: restrict memory usage of kdump kernel
Date: Tue, 28 Jul 2020 23:44:26 +1000	[thread overview]
Message-ID: <875za77o05.fsf@mpe.ellerman.id.au> (raw)
In-Reply-To: <159579231812.5790.16096865978767385505.stgit@hbathini>

Hari Bathini <hbathini@linux.ibm.com> writes:
> diff --git a/arch/powerpc/kexec/file_load_64.c b/arch/powerpc/kexec/file_load_64.c
> index 2df6f4273ddd..8df085a22fd7 100644
> --- a/arch/powerpc/kexec/file_load_64.c
> +++ b/arch/powerpc/kexec/file_load_64.c
> @@ -17,9 +17,21 @@
>  #include <linux/kexec.h>
>  #include <linux/of_fdt.h>
>  #include <linux/libfdt.h>
> +#include <linux/of_device.h>
>  #include <linux/memblock.h>
> +#include <linux/slab.h>
> +#include <asm/drmem.h>
>  #include <asm/kexec_ranges.h>
>  
> +struct umem_info {
> +	uint64_t *buf; /* data buffer for usable-memory property */
> +	uint32_t idx;  /* current index */
> +	uint32_t size; /* size allocated for the data buffer */

Use kernel types please, u64, u32.

> +	/* usable memory ranges to look up */
> +	const struct crash_mem *umrngs;

"umrngs".

Given it's part of the umem_info struct could it just be "ranges"?

> +};
> +
>  const struct kexec_file_ops * const kexec_file_loaders[] = {
>  	&kexec_elf64_ops,
>  	NULL
> @@ -74,6 +86,42 @@ static int get_exclude_memory_ranges(struct crash_mem **mem_ranges)
>  	return ret;
>  }
>  
> +/**
> + * get_usable_memory_ranges - Get usable memory ranges. This list includes
> + *                            regions like crashkernel, opal/rtas & tce-table,
> + *                            that kdump kernel could use.
> + * @mem_ranges:               Range list to add the memory ranges to.
> + *
> + * Returns 0 on success, negative errno on error.
> + */
> +static int get_usable_memory_ranges(struct crash_mem **mem_ranges)
> +{
> +	int ret;
> +
> +	/*
> +	 * prom code doesn't take kindly to missing low memory. So, add

I don't know what that's referring to, "prom code" is too vague.

> +	 * [0, crashk_res.end] instead of [crashk_res.start, crashk_res.end]
> +	 * to keep it happy.
> +	 */
> +	ret = add_mem_range(mem_ranges, 0, crashk_res.end + 1);
> +	if (ret)
> +		goto out;
> +
> +	ret = add_rtas_mem_range(mem_ranges);
> +	if (ret)
> +		goto out;
> +
> +	ret = add_opal_mem_range(mem_ranges);
> +	if (ret)
> +		goto out;
> +
> +	ret = add_tce_mem_ranges(mem_ranges);
> +out:
> +	if (ret)
> +		pr_err("Failed to setup usable memory ranges\n");
> +	return ret;
> +}
> +
>  /**
>   * __locate_mem_hole_top_down - Looks top down for a large enough memory hole
>   *                              in the memory regions between buf_min & buf_max
> @@ -273,6 +321,382 @@ static int locate_mem_hole_bottom_up_ppc64(struct kexec_buf *kbuf,
>  	return ret;
>  }
>  
> +/**
> + * check_realloc_usable_mem - Reallocate buffer if it can't accommodate entries
> + * @um_info:                  Usable memory buffer and ranges info.
> + * @cnt:                      No. of entries to accommodate.
> + *
> + * Frees up the old buffer if memory reallocation fails.
> + *
> + * Returns buffer on success, NULL on error.
> + */
> +static uint64_t *check_realloc_usable_mem(struct umem_info *um_info, int cnt)
> +{
> +	void *tbuf;
> +
> +	if (um_info->size >=
> +	    ((um_info->idx + cnt) * sizeof(*(um_info->buf))))
> +		return um_info->buf;

This is awkward.

AFAICS you only use um_info->size here, so instead why not store the
number of u64s you have space for, as num for example.

Then the above comparison becomes:

	if (um_info->num >= (um_info->idx + count))

Then you only have to calculate the size internally here for the
realloc.

> +
> +	um_info->size += MEM_RANGE_CHUNK_SZ;

	new_size = um_info->size + MEM_RANGE_CHUNK_SZ;
	tbuf = krealloc(um_info->buf, new_size, GFP_KERNEL);

> +	tbuf = krealloc(um_info->buf, um_info->size, GFP_KERNEL);
> +	if (!tbuf) {
> +		um_info->size -= MEM_RANGE_CHUNK_SZ;

Then you can drop this.

> +		return NULL;
> +	}

	um_info->size = new_size;

> +
> +	memset(tbuf + um_info->idx, 0, MEM_RANGE_CHUNK_SZ);

Just pass __GFP_ZERO to krealloc?

> +	return tbuf;
> +}
> +
> +/**
> + * add_usable_mem - Add the usable memory ranges within the given memory range
> + *                  to the buffer
> + * @um_info:        Usable memory buffer and ranges info.
> + * @base:           Base address of memory range to look for.
> + * @end:            End address of memory range to look for.
> + * @cnt:            No. of usable memory ranges added to buffer.

One caller never uses this AFAICS.

Couldn't the other caller just compare the um_info->idx before and after
the call, and avoid another pass by reference parameter.

> + *
> + * Returns 0 on success, negative errno on error.
> + */
> +static int add_usable_mem(struct umem_info *um_info, uint64_t base,
> +			  uint64_t end, int *cnt)
> +{
> +	uint64_t loc_base, loc_end, *buf;
> +	const struct crash_mem *umrngs;
> +	int i, add;

add should be bool.

> +	*cnt = 0;
> +	umrngs = um_info->umrngs;
> +	for (i = 0; i < umrngs->nr_ranges; i++) {
> +		add = 0;
> +		loc_base = umrngs->ranges[i].start;
> +		loc_end = umrngs->ranges[i].end;
> +		if (loc_base >= base && loc_end <= end)
> +			add = 1;
> +		else if (base < loc_end && end > loc_base) {
> +			if (loc_base < base)
> +				loc_base = base;
> +			if (loc_end > end)
> +				loc_end = end;
> +			add = 1;
> +		}
> +
> +		if (add) {
> +			buf = check_realloc_usable_mem(um_info, 2);
> +			if (!buf)
> +				return -ENOMEM;
> +
> +			um_info->buf = buf;
> +			buf[um_info->idx++] = cpu_to_be64(loc_base);
> +			buf[um_info->idx++] =
> +					cpu_to_be64(loc_end - loc_base + 1);
> +			(*cnt)++;
> +		}
> +	}
> +
> +	return 0;
> +}
> +
> +/**
> + * kdump_setup_usable_lmb - This is a callback function that gets called by
> + *                          walk_drmem_lmbs for every LMB to set its
> + *                          usable memory ranges.
> + * @lmb:                    LMB info.
> + * @usm:                    linux,drconf-usable-memory property value.
> + * @data:                   Pointer to usable memory buffer and ranges info.
> + *
> + * Returns 0 on success, negative errno on error.
> + */
> +static int kdump_setup_usable_lmb(struct drmem_lmb *lmb, const __be32 **usm,
> +				  void *data)
> +{
> +	struct umem_info *um_info;
> +	uint64_t base, end, *buf;
> +	int cnt, tmp_idx, ret;
> +
> +	/*
> +	 * kdump load isn't supported on kernels already booted with
> +	 * linux,drconf-usable-memory property.
> +	 */
> +	if (*usm) {
> +		pr_err("linux,drconf-usable-memory property already exists!");
> +		return -EINVAL;
> +	}
> +
> +	um_info = data;
> +	tmp_idx = um_info->idx;
> +	buf = check_realloc_usable_mem(um_info, 1);
> +	if (!buf)
> +		return -ENOMEM;
> +
> +	um_info->idx++;
> +	um_info->buf = buf;
> +	base = lmb->base_addr;
> +	end = base + drmem_lmb_size() - 1;
> +	ret = add_usable_mem(um_info, base, end, &cnt);
> +	if (!ret)
> +		um_info->buf[tmp_idx] = cpu_to_be64(cnt);
> +
> +	return ret;
> +}
> +
> +/**
> + * get_node_path_size - Get the full path length of the given node.
> + * @dn:                 Device Node.
> + *
> + * Also, counts '\0' at the end of the path.
> + * For example, /memory@0 will be "/memory@0\0" => 10 bytes.
> + *
> + * Returns the string size of the node's full path.
> + */
> +static int get_node_path_size(struct device_node *dn)
> +{
> +	int len = 0;
> +
> +	if (!dn)
> +		return 0;
> +
> +	/* Root node */
> +	if (!(dn->parent))
> +		return 2;
> +
> +	while (dn) {
> +		len += strlen(dn->full_name) + 1;
> +		dn = dn->parent;
> +	}
> +
> +	return len;
> +}
> +
> +/**
> + * get_node_path - Get the full path of the given node.
> + * @node:          Device node.
> + *
> + * Allocates buffer for node path. The caller must free the buffer
> + * after use.
> + *
> + * Returns buffer with path on success, NULL otherwise.
> + */
> +static char *get_node_path(struct device_node *node)
> +{


As discussed this can probably be replaced with snprintf(buf, "%pOF") ?


cheers

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

  parent reply	other threads:[~2020-07-28 13:44 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-26 19:36 [RESEND PATCH v5 00/11] ppc64: enable kdump support for kexec_file_load syscall Hari Bathini
2020-07-26 19:36 ` [RESEND PATCH v5 01/11] kexec_file: allow archs to handle special regions while locating memory hole Hari Bathini
2020-07-26 19:36 ` [RESEND PATCH v5 02/11] powerpc/kexec_file: mark PPC64 specific code Hari Bathini
2020-07-26 19:37 ` [RESEND PATCH v5 03/11] powerpc/kexec_file: add helper functions for getting memory ranges Hari Bathini
2020-07-28 12:58   ` Michael Ellerman
2020-07-26 19:38 ` [RESEND PATCH v5 04/11] ppc64/kexec_file: avoid stomping memory used by special regions Hari Bathini
2020-07-26 19:38 ` [RESEND PATCH v5 05/11] powerpc/drmem: make lmb walk a bit more flexible Hari Bathini
2020-07-26 19:38 ` [RESEND PATCH v5 06/11] ppc64/kexec_file: restrict memory usage of kdump kernel Hari Bathini
2020-07-28  2:10   ` Thiago Jung Bauermann
2020-07-28 13:44   ` Michael Ellerman [this message]
2020-07-28 19:34     ` Hari Bathini
2020-07-26 19:39 ` [RESEND PATCH v5 07/11] ppc64/kexec_file: enable early kernel's OPAL calls Hari Bathini
2020-07-28  2:17   ` Thiago Jung Bauermann
2020-07-28 13:46   ` Michael Ellerman
2020-07-28 19:24     ` Hari Bathini
2020-07-29  1:15       ` Michael Ellerman
2020-07-26 19:39 ` [RESEND PATCH v5 08/11] ppc64/kexec_file: setup backup region for kdump kernel Hari Bathini
2020-07-28  2:37   ` Thiago Jung Bauermann
2020-07-28 14:11   ` Michael Ellerman
2020-07-26 19:39 ` [RESEND PATCH v5 09/11] ppc64/kexec_file: prepare elfcore header for crashing kernel Hari Bathini
2020-07-26 19:40 ` [RESEND PATCH v5 10/11] ppc64/kexec_file: add appropriate regions for memory reserve map Hari Bathini
2020-07-26 19:40 ` [RESEND PATCH v5 11/11] ppc64/kexec_file: fix kexec load failure with lack of memory hole Hari Bathini
2020-07-28  2:32 ` [RESEND PATCH v5 00/11] ppc64: enable kdump support for kexec_file_load syscall piliu
2020-07-30  6:05   ` Hari Bathini

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=875za77o05.fsf@mpe.ellerman.id.au \
    --to=mpe@ellerman.id.au \
    --cc=akpm@linux-foundation.org \
    --cc=bauerman@linux.ibm.com \
    --cc=dyoung@redhat.com \
    --cc=ebiederm@xmission.com \
    --cc=hbathini@linux.ibm.com \
    --cc=kexec@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linuxppc-dev@ozlabs.org \
    --cc=mahesh@linux.ibm.com \
    --cc=nayna@linux.ibm.com \
    --cc=piliu@redhat.com \
    --cc=ptesarik@suse.cz \
    --cc=sourabhjain@linux.ibm.com \
    --cc=vgoyal@redhat.com \
    --cc=zohar@linux.ibm.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