* Re: [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd
2021-11-24 15:31 ` [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd Alex Xu (Hello71)
@ 2021-12-03 0:49 ` Nick Terrell
2021-12-05 22:52 ` Masahiro Yamada
2021-12-17 8:51 ` Sedat Dilek
2 siblings, 0 replies; 9+ messages in thread
From: Nick Terrell @ 2021-12-03 0:49 UTC (permalink / raw)
To: Alex Xu (Hello71)
Cc: Michael Forney, Masahiro Yamada, Michal Marek, Nick Desaulniers,
Ingo Molnar, Sedat Dilek, Kees Cook, linux-kernel, linux-kbuild
> On Nov 24, 2021, at 7:31 AM, Alex Xu (Hello71) <alex_y_xu@yahoo.ca> wrote:
>
> Otherwise, it allocates 2 GB of memory at once. Even though the majority
> of this memory is never touched, the default heuristic overcommit
> refuses this request if less than 2 GB of RAM+swap is currently
> available. This results in "zstd: error 11 : Allocation error : not
> enough memory" and the kernel failing to build.
>
> When the size is specified, zstd will reduce the memory request
> appropriately. For typical kernel sizes of ~32 MB, the largest mmap
> request will be reduced to 512 MB, which will succeed on all but the
> smallest devices.
>
> For inputs around this size, --stream-size --no-content-size may
> slightly decrease the compressed size, or slightly increase it:
> https://github.com/facebook/zstd/issues/2848.
>
> Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca>
> ---
> scripts/Makefile.lib | 12 ++++++++++--
> 1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
> index ca901814986a..c98a82ca38e6 100644
> --- a/scripts/Makefile.lib
> +++ b/scripts/Makefile.lib
> @@ -466,12 +466,20 @@ quiet_cmd_xzmisc = XZMISC $@
> # single pass, so zstd doesn't need to allocate a window buffer. When streaming
> # decompression is used, like initramfs decompression, zstd22 should likely not
> # be used because it would require zstd to allocate a 128 MB buffer.
> +#
> +# --stream-size to reduce zstd memory usage (otherwise zstd -22 --ultra
> +# allocates, but does not use, 2 GB) and potentially improve compression.
> +#
> +# --no-content-size to save three bytes which we do not use (we use size_append).
> +
> +# zstd --stream-size is only supported since 1.4.4
> +zstd_stream_size = $(shell $(ZSTD) -1c --stream-size=0 --no-content-size </dev/null >/dev/null 2>&1 && printf '%s' '--stream-size=$(total_size) --no-content-size')
>
> quiet_cmd_zstd = ZSTD $@
> - cmd_zstd = { cat $(real-prereqs) | $(ZSTD) -19; $(size_append); } > $@
> + cmd_zstd = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -19; $(size_append); } > $@
>
> quiet_cmd_zstd22 = ZSTD22 $@
> - cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) -22 --ultra; $(size_append); } > $@
> + cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -22 --ultra; $(size_append); } > $@
>
> # ASM offsets
> # ---------------------------------------------------------------------------
> --
> 2.34.0
>
You can add:
Tested-by: Nick Terrell <terrelln@fb.com>
Reviewed-by: Nick Terrell <terrelln@fb.com>
Best,
Nick Terrell
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd
2021-11-24 15:31 ` [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd Alex Xu (Hello71)
2021-12-03 0:49 ` Nick Terrell
@ 2021-12-05 22:52 ` Masahiro Yamada
2021-12-06 18:42 ` Nick Terrell
2021-12-17 8:51 ` Sedat Dilek
2 siblings, 1 reply; 9+ messages in thread
From: Masahiro Yamada @ 2021-12-05 22:52 UTC (permalink / raw)
To: Alex Xu (Hello71)
Cc: Michael Forney, Michal Marek, Nick Desaulniers, Nick Terrell,
Ingo Molnar, Sedat Dilek, Kees Cook, Linux Kernel Mailing List,
Linux Kbuild mailing list
On Thu, Nov 25, 2021 at 12:30 AM Alex Xu (Hello71) <alex_y_xu@yahoo.ca> wrote:
>
> Otherwise, it allocates 2 GB of memory at once. Even though the majority
> of this memory is never touched, the default heuristic overcommit
> refuses this request if less than 2 GB of RAM+swap is currently
> available. This results in "zstd: error 11 : Allocation error : not
> enough memory" and the kernel failing to build.
>
> When the size is specified, zstd will reduce the memory request
> appropriately. For typical kernel sizes of ~32 MB, the largest mmap
> request will be reduced to 512 MB, which will succeed on all but the
> smallest devices.
>
> For inputs around this size, --stream-size --no-content-size may
> slightly decrease the compressed size, or slightly increase it:
> https://github.com/facebook/zstd/issues/2848.
>
> Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca>
The reason why we need this workaround is just because we do
"cat and compress". zstd must allocate a huge memory beforehand
since it cannot predict how long the stream it will receive.
If zstd is given with a file name, it can fstat it to know its file size
and allocate the minimal amount of memory.
This is my test.
I used 'ulimit' to set the upper limit of the memory the zstd can use.
[test steps]
# Create a 1kB file
$ truncate --size=1k dummy
# Set the memory size limit to 10MB
$ ulimit -S -v 10240
# Pass the file as a argument; success
$ zstd -19 -o dummy.zst dummy
dummy : 2.15% ( 1024 => 22 bytes, dummy.zst)
# cat and zstd; fail
$ cat dummy | zstd -19 > dummy.zst
zstd: error 11 : Allocation error : not enough memory
# cat and zstd --stream-size; success
$ cat dummy | zstd -19 --stream-size=1024 > dummy.zst
scripts/Makefile.modinst was written in such a way
that zstd can know the file size by itself.
cmd_zstd = $(ZSTD) -T0 --rm -f -q $<
We cannot rewrite scripts/Makefile.lib in that way because
arch/x86/boot/compress/Makefile concatenates two files before
compression. And this is the only use-case of this feature.
So, I am seriously considering to revert this commit:
commit d3dd3b5a29bb9582957451531fed461628dfc834
Author: H. Peter Anvin <hpa@zytor.com>
Date: Tue May 5 21:17:15 2009 -0700
kbuild: allow compressors (gzip, bzip2, lzma) to take multiple inputs
With that commit reverted, zstd will take a single input file,
and we can do "zstd -o <output> <input>".
So, I will take some time to investigate that approach.
> ---
> scripts/Makefile.lib | 12 ++++++++++--
> 1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
> index ca901814986a..c98a82ca38e6 100644
> --- a/scripts/Makefile.lib
> +++ b/scripts/Makefile.lib
> @@ -466,12 +466,20 @@ quiet_cmd_xzmisc = XZMISC $@
> # single pass, so zstd doesn't need to allocate a window buffer. When streaming
> # decompression is used, like initramfs decompression, zstd22 should likely not
> # be used because it would require zstd to allocate a 128 MB buffer.
> +#
> +# --stream-size to reduce zstd memory usage (otherwise zstd -22 --ultra
> +# allocates, but does not use, 2 GB) and potentially improve compression.
> +#
> +# --no-content-size to save three bytes which we do not use (we use size_append).
> +
> +# zstd --stream-size is only supported since 1.4.4
> +zstd_stream_size = $(shell $(ZSTD) -1c --stream-size=0 --no-content-size </dev/null >/dev/null 2>&1 && printf '%s' '--stream-size=$(total_size) --no-content-size')
>
> quiet_cmd_zstd = ZSTD $@
> - cmd_zstd = { cat $(real-prereqs) | $(ZSTD) -19; $(size_append); } > $@
> + cmd_zstd = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -19; $(size_append); } > $@
>
> quiet_cmd_zstd22 = ZSTD22 $@
> - cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) -22 --ultra; $(size_append); } > $@
> + cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -22 --ultra; $(size_append); } > $@
>
> # ASM offsets
> # ---------------------------------------------------------------------------
> --
> 2.34.0
>
--
Best Regards
Masahiro Yamada
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd
2021-12-05 22:52 ` Masahiro Yamada
@ 2021-12-06 18:42 ` Nick Terrell
0 siblings, 0 replies; 9+ messages in thread
From: Nick Terrell @ 2021-12-06 18:42 UTC (permalink / raw)
To: Masahiro Yamada
Cc: Alex Xu (Hello71),
Michael Forney, Michal Marek, Nick Desaulniers, Ingo Molnar,
Sedat Dilek, Kees Cook, Linux Kernel Mailing List,
Linux Kbuild mailing list
> On Dec 5, 2021, at 2:52 PM, Masahiro Yamada <masahiroy@kernel.org> wrote:
>
> On Thu, Nov 25, 2021 at 12:30 AM Alex Xu (Hello71) <alex_y_xu@yahoo.ca> wrote:
>>
>> Otherwise, it allocates 2 GB of memory at once. Even though the majority
>> of this memory is never touched, the default heuristic overcommit
>> refuses this request if less than 2 GB of RAM+swap is currently
>> available. This results in "zstd: error 11 : Allocation error : not
>> enough memory" and the kernel failing to build.
>>
>> When the size is specified, zstd will reduce the memory request
>> appropriately. For typical kernel sizes of ~32 MB, the largest mmap
>> request will be reduced to 512 MB, which will succeed on all but the
>> smallest devices.
>>
>> For inputs around this size, --stream-size --no-content-size may
>> slightly decrease the compressed size, or slightly increase it:
>> https://github.com/facebook/zstd/issues/2848.
>>
>> Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca>
>
>
>
>
> The reason why we need this workaround is just because we do
> "cat and compress". zstd must allocate a huge memory beforehand
> since it cannot predict how long the stream it will receive.
>
> If zstd is given with a file name, it can fstat it to know its file size
> and allocate the minimal amount of memory.
>
>
> This is my test.
> I used 'ulimit' to set the upper limit of the memory the zstd can use.
>
>
> [test steps]
>
> # Create a 1kB file
> $ truncate --size=1k dummy
>
> # Set the memory size limit to 10MB
> $ ulimit -S -v 10240
>
> # Pass the file as a argument; success
> $ zstd -19 -o dummy.zst dummy
> dummy : 2.15% ( 1024 => 22 bytes, dummy.zst)
>
> # cat and zstd; fail
> $ cat dummy | zstd -19 > dummy.zst
> zstd: error 11 : Allocation error : not enough memory
>
> # cat and zstd --stream-size; success
> $ cat dummy | zstd -19 --stream-size=1024 > dummy.zst
>
>
>
>
> scripts/Makefile.modinst was written in such a way
> that zstd can know the file size by itself.
>
> cmd_zstd = $(ZSTD) -T0 --rm -f -q $<
>
>
> We cannot rewrite scripts/Makefile.lib in that way because
> arch/x86/boot/compress/Makefile concatenates two files before
> compression. And this is the only use-case of this feature.
>
> So, I am seriously considering to revert this commit:
>
> commit d3dd3b5a29bb9582957451531fed461628dfc834
> Author: H. Peter Anvin <hpa@zytor.com>
> Date: Tue May 5 21:17:15 2009 -0700
>
> kbuild: allow compressors (gzip, bzip2, lzma) to take multiple inputs
>
>
>
>
> With that commit reverted, zstd will take a single input file,
> and we can do "zstd -o <output> <input>".
>
>
> So, I will take some time to investigate that approach.
This will definitely work from a zstd perspective. All versions
of zstd will downsize their memory usage to match the file size.
Best,
Nick Terrell
>> ---
>> scripts/Makefile.lib | 12 ++++++++++--
>> 1 file changed, 10 insertions(+), 2 deletions(-)
>>
>> diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
>> index ca901814986a..c98a82ca38e6 100644
>> --- a/scripts/Makefile.lib
>> +++ b/scripts/Makefile.lib
>> @@ -466,12 +466,20 @@ quiet_cmd_xzmisc = XZMISC $@
>> # single pass, so zstd doesn't need to allocate a window buffer. When streaming
>> # decompression is used, like initramfs decompression, zstd22 should likely not
>> # be used because it would require zstd to allocate a 128 MB buffer.
>> +#
>> +# --stream-size to reduce zstd memory usage (otherwise zstd -22 --ultra
>> +# allocates, but does not use, 2 GB) and potentially improve compression.
>> +#
>> +# --no-content-size to save three bytes which we do not use (we use size_append).
>> +
>> +# zstd --stream-size is only supported since 1.4.4
>> +zstd_stream_size = $(shell $(ZSTD) -1c --stream-size=0 --no-content-size </dev/null >/dev/null 2>&1 && printf '%s' '--stream-size=$(total_size) --no-content-size')
>>
>> quiet_cmd_zstd = ZSTD $@
>> - cmd_zstd = { cat $(real-prereqs) | $(ZSTD) -19; $(size_append); } > $@
>> + cmd_zstd = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -19; $(size_append); } > $@
>>
>> quiet_cmd_zstd22 = ZSTD22 $@
>> - cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) -22 --ultra; $(size_append); } > $@
>> + cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -22 --ultra; $(size_append); } > $@
>>
>> # ASM offsets
>> # ---------------------------------------------------------------------------
>> --
>> 2.34.0
>>
>
>
> --
> Best Regards
> Masahiro Yamada
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd
2021-11-24 15:31 ` [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd Alex Xu (Hello71)
2021-12-03 0:49 ` Nick Terrell
2021-12-05 22:52 ` Masahiro Yamada
@ 2021-12-17 8:51 ` Sedat Dilek
2021-12-17 13:44 ` Sedat Dilek
2 siblings, 1 reply; 9+ messages in thread
From: Sedat Dilek @ 2021-12-17 8:51 UTC (permalink / raw)
To: Alex Xu (Hello71), Nick Terrell
Cc: Michael Forney, Masahiro Yamada, Michal Marek, Nick Desaulniers,
Ingo Molnar, Kees Cook, linux-kernel, linux-kbuild
On Wed, Nov 24, 2021 at 4:30 PM Alex Xu (Hello71) <alex_y_xu@yahoo.ca> wrote:
>
> Otherwise, it allocates 2 GB of memory at once. Even though the majority
> of this memory is never touched, the default heuristic overcommit
> refuses this request if less than 2 GB of RAM+swap is currently
> available. This results in "zstd: error 11 : Allocation error : not
> enough memory" and the kernel failing to build.
>
> When the size is specified, zstd will reduce the memory request
> appropriately. For typical kernel sizes of ~32 MB, the largest mmap
> request will be reduced to 512 MB, which will succeed on all but the
> smallest devices.
>
> For inputs around this size, --stream-size --no-content-size may
> slightly decrease the compressed size, or slightly increase it:
> https://github.com/facebook/zstd/issues/2848.
>
Hi Alex and Nick T.,
some questions:
Can I apply this patch as a single patch - without patch 1/2?
Is there an impact also on the kernel's ZRAM/ZSWAP support plus using
ZSTD as (de)comp-algo?
Here I have:
$ grep -i zstd /boot/config-5.15.7-1-amd64-clang13-lto | egrep -i 'zram|zswap'
CONFIG_ZSWAP_COMPRESSOR_DEFAULT_ZSTD=y
CONFIG_ZSWAP_COMPRESSOR_DEFAULT="zstd"
CONFIG_ZRAM_DEF_COMP_ZSTD=y
CONFIG_ZRAM_DEF_COMP="zstd"
Thanks.
Regards,
- Sedat -
> Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca>
> ---
> scripts/Makefile.lib | 12 ++++++++++--
> 1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
> index ca901814986a..c98a82ca38e6 100644
> --- a/scripts/Makefile.lib
> +++ b/scripts/Makefile.lib
> @@ -466,12 +466,20 @@ quiet_cmd_xzmisc = XZMISC $@
> # single pass, so zstd doesn't need to allocate a window buffer. When streaming
> # decompression is used, like initramfs decompression, zstd22 should likely not
> # be used because it would require zstd to allocate a 128 MB buffer.
> +#
> +# --stream-size to reduce zstd memory usage (otherwise zstd -22 --ultra
> +# allocates, but does not use, 2 GB) and potentially improve compression.
> +#
> +# --no-content-size to save three bytes which we do not use (we use size_append).
> +
> +# zstd --stream-size is only supported since 1.4.4
> +zstd_stream_size = $(shell $(ZSTD) -1c --stream-size=0 --no-content-size </dev/null >/dev/null 2>&1 && printf '%s' '--stream-size=$(total_size) --no-content-size')
>
> quiet_cmd_zstd = ZSTD $@
> - cmd_zstd = { cat $(real-prereqs) | $(ZSTD) -19; $(size_append); } > $@
> + cmd_zstd = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -19; $(size_append); } > $@
>
> quiet_cmd_zstd22 = ZSTD22 $@
> - cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) -22 --ultra; $(size_append); } > $@
> + cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -22 --ultra; $(size_append); } > $@
>
> # ASM offsets
> # ---------------------------------------------------------------------------
> --
> 2.34.0
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v2 2/2] kbuild: pass --stream-size --no-content-size to zstd
2021-12-17 8:51 ` Sedat Dilek
@ 2021-12-17 13:44 ` Sedat Dilek
0 siblings, 0 replies; 9+ messages in thread
From: Sedat Dilek @ 2021-12-17 13:44 UTC (permalink / raw)
To: Alex Xu (Hello71), Nick Terrell
Cc: Michael Forney, Masahiro Yamada, Michal Marek, Nick Desaulniers,
Ingo Molnar, Kees Cook, linux-kernel, linux-kbuild
On Fri, Dec 17, 2021 at 9:51 AM Sedat Dilek <sedat.dilek@gmail.com> wrote:
>
> On Wed, Nov 24, 2021 at 4:30 PM Alex Xu (Hello71) <alex_y_xu@yahoo.ca> wrote:
> >
> > Otherwise, it allocates 2 GB of memory at once. Even though the majority
> > of this memory is never touched, the default heuristic overcommit
> > refuses this request if less than 2 GB of RAM+swap is currently
> > available. This results in "zstd: error 11 : Allocation error : not
> > enough memory" and the kernel failing to build.
> >
> > When the size is specified, zstd will reduce the memory request
> > appropriately. For typical kernel sizes of ~32 MB, the largest mmap
> > request will be reduced to 512 MB, which will succeed on all but the
> > smallest devices.
> >
> > For inputs around this size, --stream-size --no-content-size may
> > slightly decrease the compressed size, or slightly increase it:
> > https://github.com/facebook/zstd/issues/2848.
> >
>
> Hi Alex and Nick T.,
>
> some questions:
>
> Can I apply this patch as a single patch - without patch 1/2?
>
> Is there an impact also on the kernel's ZRAM/ZSWAP support plus using
> ZSTD as (de)comp-algo?
>
> Here I have:
>
> $ grep -i zstd /boot/config-5.15.7-1-amd64-clang13-lto | egrep -i 'zram|zswap'
> CONFIG_ZSWAP_COMPRESSOR_DEFAULT_ZSTD=y
> CONFIG_ZSWAP_COMPRESSOR_DEFAULT="zstd"
> CONFIG_ZRAM_DEF_COMP_ZSTD=y
> CONFIG_ZRAM_DEF_COMP="zstd"
>
$ egrep 'stream-size' build-log_5.15.9-1-amd64-clang13-lto.txt
49360: { cat arch/x86/boot/compressed/vmlinux.bin
arch/x86/boot/compressed/vmlinux.relocs | zstd --stream-size=53340760
--no-content-size -22 --ultra; printf \130\352
\055\003; } > arch/x86/boot/compressed/vmlinux.bin.zst
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
- Sedat -
> Thanks.
>
> Regards,
> - Sedat -
>
> > Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca>
> > ---
> > scripts/Makefile.lib | 12 ++++++++++--
> > 1 file changed, 10 insertions(+), 2 deletions(-)
> >
> > diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
> > index ca901814986a..c98a82ca38e6 100644
> > --- a/scripts/Makefile.lib
> > +++ b/scripts/Makefile.lib
> > @@ -466,12 +466,20 @@ quiet_cmd_xzmisc = XZMISC $@
> > # single pass, so zstd doesn't need to allocate a window buffer. When streaming
> > # decompression is used, like initramfs decompression, zstd22 should likely not
> > # be used because it would require zstd to allocate a 128 MB buffer.
> > +#
> > +# --stream-size to reduce zstd memory usage (otherwise zstd -22 --ultra
> > +# allocates, but does not use, 2 GB) and potentially improve compression.
> > +#
> > +# --no-content-size to save three bytes which we do not use (we use size_append).
> > +
> > +# zstd --stream-size is only supported since 1.4.4
> > +zstd_stream_size = $(shell $(ZSTD) -1c --stream-size=0 --no-content-size </dev/null >/dev/null 2>&1 && printf '%s' '--stream-size=$(total_size) --no-content-size')
> >
> > quiet_cmd_zstd = ZSTD $@
> > - cmd_zstd = { cat $(real-prereqs) | $(ZSTD) -19; $(size_append); } > $@
> > + cmd_zstd = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -19; $(size_append); } > $@
> >
> > quiet_cmd_zstd22 = ZSTD22 $@
> > - cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) -22 --ultra; $(size_append); } > $@
> > + cmd_zstd22 = { cat $(real-prereqs) | $(ZSTD) $(zstd_stream_size) -22 --ultra; $(size_append); } > $@
> >
> > # ASM offsets
> > # ---------------------------------------------------------------------------
> > --
> > 2.34.0
> >
^ permalink raw reply [flat|nested] 9+ messages in thread