2024-02-20 02:21:33

by yang.zhang

Subject: [PATCH v2] kexec: copy only happens before uchunk goes to zero

From: "yang.zhang" <[email protected]>

When load segments, all of the copying and the rest
only happens before uchunk goes to zero.

Signed-off-by: yang.zhang <[email protected]>

---
v1 -> v2:
- Only copy before uchunk goes to zero

V1: https://lore.kernel.org/lkml/[email protected]/
---
kernel/kexec_core.c | 44 ++++++++++++++++++++++++--------------------
1 file changed, 24 insertions(+), 20 deletions(-)

diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c
index d08fc7b5db97..2fc3d0e3715a 100644
--- a/kernel/kexec_core.c
+++ b/kernel/kexec_core.c
@@ -800,22 +800,24 @@ static int kimage_load_normal_segment(struct kimage *image,
PAGE_SIZE - (maddr & ~PAGE_MASK));
uchunk = min(ubytes, mchunk);

- /* For file based kexec, source pages are in kernel memory */
- if (image->file_mode)
- memcpy(ptr, kbuf, uchunk);
- else
- result = copy_from_user(ptr, buf, uchunk);
+ if (uchunk) {
+ /* For file based kexec, source pages are in kernel memory */
+ if (image->file_mode)
+ memcpy(ptr, kbuf, uchunk);
+ else
+ result = copy_from_user(ptr, buf, uchunk);
+ ubytes -= uchunk;
+ if (image->file_mode)
+ kbuf += uchunk;
+ else
+ buf += uchunk;
+ }
kunmap_local(ptr);
if (result) {
result = -EFAULT;
goto out;
}
- ubytes -= uchunk;
maddr += mchunk;
- if (image->file_mode)
- kbuf += mchunk;
- else
- buf += mchunk;
mbytes -= mchunk;

cond_resched();
@@ -866,11 +868,18 @@ static int kimage_load_crash_segment(struct kimage *image,
memset(ptr + uchunk, 0, mchunk - uchunk);
}

- /* For file based kexec, source pages are in kernel memory */
- if (image->file_mode)
- memcpy(ptr, kbuf, uchunk);
- else
- result = copy_from_user(ptr, buf, uchunk);
+ if (uchunk) {
+ /* For file based kexec, source pages are in kernel memory */
+ if (image->file_mode)
+ memcpy(ptr, kbuf, uchunk);
+ else
+ result = copy_from_user(ptr, buf, uchunk);
+ ubytes -= uchunk;
+ if (image->file_mode)
+ kbuf += uchunk;
+ else
+ buf += uchunk;
+ }
kexec_flush_icache_page(page);
kunmap_local(ptr);
arch_kexec_pre_free_pages(page_address(page), 1);
@@ -878,12 +887,7 @@ static int kimage_load_crash_segment(struct kimage *image,
result = -EFAULT;
goto out;
}
- ubytes -= uchunk;
maddr += mchunk;
- if (image->file_mode)
- kbuf += mchunk;
- else
- buf += mchunk;
mbytes -= mchunk;

cond_resched();
--
2.25.1

2024-02-22 07:46:02

by Baoquan He

Subject: Re: [PATCH v2] kexec: copy only happens before uchunk goes to zero

On 02/20/24 at 10:11am, yang.zhang wrote:
> From: "yang.zhang" <[email protected]>
>
> When load segments, all of the copying and the rest
> only happens before uchunk goes to zero.

The code change looks good, while the log can be improved with more
details so that people can more easily get what's wrong and fixed.

===
When loading segments, ubytes is <= mbytes. When ubytes is exhausted,
there could be remaining mbytes. Then in the while loop, the buf pointer
advancing with mchunk will causing meaningless reading even though it
doesn't harm.

So let's change to make sure that all of the copying and the rest only
happens before uchunk goes to zero.
===

Please take above words as reference to udpate patch log and post v3,
and you can add:

Acked-by: Baoquan He <[email protected]>

>
> Signed-off-by: yang.zhang <[email protected]>
>
> ---
> v1 -> v2:
> - Only copy before uchunk goes to zero
>
> V1: https://lore.kernel.org/lkml/[email protected]/
> ---
> kernel/kexec_core.c | 44 ++++++++++++++++++++++++--------------------
> 1 file changed, 24 insertions(+), 20 deletions(-)
>
> diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c
> index d08fc7b5db97..2fc3d0e3715a 100644
> --- a/kernel/kexec_core.c
> +++ b/kernel/kexec_core.c
> @@ -800,22 +800,24 @@ static int kimage_load_normal_segment(struct kimage *image,
> PAGE_SIZE - (maddr & ~PAGE_MASK));
> uchunk = min(ubytes, mchunk);
>
> - /* For file based kexec, source pages are in kernel memory */
> - if (image->file_mode)
> - memcpy(ptr, kbuf, uchunk);
> - else
> - result = copy_from_user(ptr, buf, uchunk);
> + if (uchunk) {
> + /* For file based kexec, source pages are in kernel memory */
> + if (image->file_mode)
> + memcpy(ptr, kbuf, uchunk);
> + else
> + result = copy_from_user(ptr, buf, uchunk);
> + ubytes -= uchunk;
> + if (image->file_mode)
> + kbuf += uchunk;
> + else
> + buf += uchunk;
> + }
> kunmap_local(ptr);
> if (result) {
> result = -EFAULT;
> goto out;
> }
> - ubytes -= uchunk;
> maddr += mchunk;
> - if (image->file_mode)
> - kbuf += mchunk;
> - else
> - buf += mchunk;
> mbytes -= mchunk;
>
> cond_resched();
> @@ -866,11 +868,18 @@ static int kimage_load_crash_segment(struct kimage *image,
> memset(ptr + uchunk, 0, mchunk - uchunk);
> }
>
> - /* For file based kexec, source pages are in kernel memory */
> - if (image->file_mode)
> - memcpy(ptr, kbuf, uchunk);
> - else
> - result = copy_from_user(ptr, buf, uchunk);
> + if (uchunk) {
> + /* For file based kexec, source pages are in kernel memory */
> + if (image->file_mode)
> + memcpy(ptr, kbuf, uchunk);
> + else
> + result = copy_from_user(ptr, buf, uchunk);
> + ubytes -= uchunk;
> + if (image->file_mode)
> + kbuf += uchunk;
> + else
> + buf += uchunk;
> + }
> kexec_flush_icache_page(page);
> kunmap_local(ptr);
> arch_kexec_pre_free_pages(page_address(page), 1);
> @@ -878,12 +887,7 @@ static int kimage_load_crash_segment(struct kimage *image,
> result = -EFAULT;
> goto out;
> }
> - ubytes -= uchunk;
> maddr += mchunk;
> - if (image->file_mode)
> - kbuf += mchunk;
> - else
> - buf += mchunk;
> mbytes -= mchunk;
>
> cond_resched();
> --
> 2.25.1
>

2024-02-22 09:24:50

by yang.zhang

Subject: Re:Re: [PATCH v2] kexec: copy only happens before uchunk goes to zero

At 2024-02-22 15:45:38, "Baoquan He" <[email protected]> wrote:
>On 02/20/24 at 10:11am, yang.zhang wrote:
>> From: "yang.zhang" <[email protected]>
>>
>> When load segments, all of the copying and the rest
>> only happens before uchunk goes to zero.
>
>The code change looks good, while the log can be improved with more
>details so that people can more easily get what's wrong and fixed.
>
>===
>When loading segments, ubytes is <= mbytes. When ubytes is exhausted,
>there could be remaining mbytes. Then in the while loop, the buf pointer
>advancing with mchunk will causing meaningless reading even though it
>doesn't harm.
>
>So let's change to make sure that all of the copying and the rest only
>happens before uchunk goes to zero.
>===
>
>Please take above words as reference to udpate patch log and post v3,
>and you can add:
>

>Acked-by: Baoquan He <[email protected]>
Thanks, i update the commits and post v3.>
>>
>> Signed-off-by: yang.zhang <[email protected]>
>>
>> ---
>> v1 -> v2:
>> - Only copy before uchunk goes to zero
>>
>> V1: https://lore.kernel.org/lkml/[email protected]/
>> ---
>> kernel/kexec_core.c | 44 ++++++++++++++++++++++++--------------------
>> 1 file changed, 24 insertions(+), 20 deletions(-)
>>
>> diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c
>> index d08fc7b5db97..2fc3d0e3715a 100644
>> --- a/kernel/kexec_core.c
>> +++ b/kernel/kexec_core.c
>> @@ -800,22 +800,24 @@ static int kimage_load_normal_segment(struct kimage *image,
>> PAGE_SIZE - (maddr & ~PAGE_MASK));
>> uchunk = min(ubytes, mchunk);
>>
>> - /* For file based kexec, source pages are in kernel memory */
>> - if (image->file_mode)
>> - memcpy(ptr, kbuf, uchunk);
>> - else
>> - result = copy_from_user(ptr, buf, uchunk);
>> + if (uchunk) {
>> + /* For file based kexec, source pages are in kernel memory */
>> + if (image->file_mode)
>> + memcpy(ptr, kbuf, uchunk);
>> + else
>> + result = copy_from_user(ptr, buf, uchunk);
>> + ubytes -= uchunk;
>> + if (image->file_mode)
>> + kbuf += uchunk;
>> + else
>> + buf += uchunk;
>> + }
>> kunmap_local(ptr);
>> if (result) {
>> result = -EFAULT;
>> goto out;
>> }
>> - ubytes -= uchunk;
>> maddr += mchunk;
>> - if (image->file_mode)
>> - kbuf += mchunk;
>> - else
>> - buf += mchunk;
>> mbytes -= mchunk;
>>
>> cond_resched();
>> @@ -866,11 +868,18 @@ static int kimage_load_crash_segment(struct kimage *image,
>> memset(ptr + uchunk, 0, mchunk - uchunk);
>> }
>>
>> - /* For file based kexec, source pages are in kernel memory */
>> - if (image->file_mode)
>> - memcpy(ptr, kbuf, uchunk);
>> - else
>> - result = copy_from_user(ptr, buf, uchunk);
>> + if (uchunk) {
>> + /* For file based kexec, source pages are in kernel memory */
>> + if (image->file_mode)
>> + memcpy(ptr, kbuf, uchunk);
>> + else
>> + result = copy_from_user(ptr, buf, uchunk);
>> + ubytes -= uchunk;
>> + if (image->file_mode)
>> + kbuf += uchunk;
>> + else
>> + buf += uchunk;
>> + }
>> kexec_flush_icache_page(page);
>> kunmap_local(ptr);
>> arch_kexec_pre_free_pages(page_address(page), 1);
>> @@ -878,12 +887,7 @@ static int kimage_load_crash_segment(struct kimage *image,
>> result = -EFAULT;
>> goto out;
>> }
>> - ubytes -= uchunk;
>> maddr += mchunk;
>> - if (image->file_mode)
>> - kbuf += mchunk;
>> - else
>> - buf += mchunk;
>> mbytes -= mchunk;
>>
>> cond_resched();
>> --
>> 2.25.1
>>