* [Buildroot] [PATCH 1/8] arch: introduce minimal required gcc version
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-10-07 9:59 ` Romain Naour
2017-09-03 9:44 ` [Buildroot] [PATCH 2/8] package/gcc: hide versions too old for the current arch Yann E. MORIN
` (8 subsequent siblings)
9 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Some CPU variants require that a recent-enough gcc be selected. For
example, ARM's cortex-a35 requires gcc-5, while cortex-a73 requires
gcc-7. Same goes for other architectures, of course.
Currently, we hard-code every such conditions in the gcc version choice,
as well as in the individual external toolchains.
However, as we add even more CPU variants, the conditions are getting
more and more complex to write and maintain.
Introduce new symbols, that architectures can select if they have a
specific requirement on the gcc version. gcc and external toolchains
can then properly depend on those symbols.
The burden of maintaining the requirements on the gcc version now falls
down to the architeture, instead of being split up in gcc and all the
external toolchains.
As the oldest gcc version to handle, we can either choose gcc-4.9, as
the odlest version we support in our internal toolchain, or choose
gcc-4.8, as the oldest external toolchain we support (except for the
custom ones, but they'll be handled specifically in upcoming changes).
We choose to go back up to gcc-4.8.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/arch/Config.in b/arch/Config.in
index c10bf16809..607da36c1d 100644
--- a/arch/Config.in
+++ b/arch/Config.in
@@ -261,6 +261,27 @@ config BR2_ARCH_HAS_TOOLCHAIN_BUILDROOT
bool
default y if !BR2_ARCH_HAS_NO_TOOLCHAIN_BUILDROOT
+# The following symbols are selected by the individual
+# Config.in.$ARCH files
+config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
+ bool
+
+config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
+ bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
+
+config BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+ bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
+
+config BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+ bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+
+config BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+ bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+
# The following string values are defined by the individual
# Config.in.$ARCH files
config BR2_ARCH
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 1/8] arch: introduce minimal required gcc version
2017-09-03 9:44 ` [Buildroot] [PATCH 1/8] arch: introduce minimal required " Yann E. MORIN
@ 2017-10-07 9:59 ` Romain Naour
2017-10-07 10:26 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Romain Naour @ 2017-10-07 9:59 UTC (permalink / raw)
To: buildroot
Yann, All,
Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
> Some CPU variants require that a recent-enough gcc be selected. For
> example, ARM's cortex-a35 requires gcc-5, while cortex-a73 requires
> gcc-7. Same goes for other architectures, of course.
>
> Currently, we hard-code every such conditions in the gcc version choice,
> as well as in the individual external toolchains.
>
> However, as we add even more CPU variants, the conditions are getting
> more and more complex to write and maintain.
>
> Introduce new symbols, that architectures can select if they have a
> specific requirement on the gcc version. gcc and external toolchains
> can then properly depend on those symbols.
>
> The burden of maintaining the requirements on the gcc version now falls
> down to the architeture, instead of being split up in gcc and all the
> external toolchains.
>
> As the oldest gcc version to handle, we can either choose gcc-4.9, as
> the odlest version we support in our internal toolchain, or choose
> gcc-4.8, as the oldest external toolchain we support (except for the
> custom ones, but they'll be handled specifically in upcoming changes).
> We choose to go back up to gcc-4.8.
>
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> ---
> arch/Config.in | 21 +++++++++++++++++++++
> 1 file changed, 21 insertions(+)
>
> diff --git a/arch/Config.in b/arch/Config.in
> index c10bf16809..607da36c1d 100644
> --- a/arch/Config.in
> +++ b/arch/Config.in
> @@ -261,6 +261,27 @@ config BR2_ARCH_HAS_TOOLCHAIN_BUILDROOT
> bool
> default y if !BR2_ARCH_HAS_NO_TOOLCHAIN_BUILDROOT
>
> +# The following symbols are selected by the individual
> +# Config.in.$ARCH files
I believe something is missing from this patch.
Best regards,
Romain
> +config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> + bool
> +
> +config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> + bool
> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> +
> +config BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> + bool
> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> +
> +config BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> + bool
> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> +
> +config BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> + bool
> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> +
> # The following string values are defined by the individual
> # Config.in.$ARCH files
> config BR2_ARCH
>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 1/8] arch: introduce minimal required gcc version
2017-10-07 9:59 ` Romain Naour
@ 2017-10-07 10:26 ` Yann E. MORIN
0 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-10-07 10:26 UTC (permalink / raw)
To: buildroot
Romain, All,
On 2017-10-07 11:59 +0200, Romain Naour spake thusly:
> Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
> > Some CPU variants require that a recent-enough gcc be selected. For
> > example, ARM's cortex-a35 requires gcc-5, while cortex-a73 requires
> > gcc-7. Same goes for other architectures, of course.
> >
> > Currently, we hard-code every such conditions in the gcc version choice,
> > as well as in the individual external toolchains.
> >
> > However, as we add even more CPU variants, the conditions are getting
> > more and more complex to write and maintain.
> >
> > Introduce new symbols, that architectures can select if they have a
> > specific requirement on the gcc version. gcc and external toolchains
> > can then properly depend on those symbols.
> >
> > The burden of maintaining the requirements on the gcc version now falls
> > down to the architeture, instead of being split up in gcc and all the
> > external toolchains.
> >
> > As the oldest gcc version to handle, we can either choose gcc-4.9, as
> > the odlest version we support in our internal toolchain, or choose
> > gcc-4.8, as the oldest external toolchain we support (except for the
> > custom ones, but they'll be handled specifically in upcoming changes).
> > We choose to go back up to gcc-4.8.
> >
> > Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> > Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> > ---
> > arch/Config.in | 21 +++++++++++++++++++++
> > 1 file changed, 21 insertions(+)
> >
> > diff --git a/arch/Config.in b/arch/Config.in
> > index c10bf16809..607da36c1d 100644
> > --- a/arch/Config.in
> > +++ b/arch/Config.in
> > @@ -261,6 +261,27 @@ config BR2_ARCH_HAS_TOOLCHAIN_BUILDROOT
> > bool
> > default y if !BR2_ARCH_HAS_NO_TOOLCHAIN_BUILDROOT
> >
> > +# The following symbols are selected by the individual
> > +# Config.in.$ARCH files
>
> I believe something is missing from this patch.
Nope, this patch only adds the symbols. The follow-up patches convert
the individual architectures, one by one, over to using them.
Regards,
Yann E. MORIN.
> Best regards,
> Romain
>
> > +config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > + bool
> > +
> > +config BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > + bool
> > + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > +
> > +config BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > + bool
> > + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > +
> > +config BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > + bool
> > + select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > +
> > +config BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > + bool
> > + select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > +
> > # The following string values are defined by the individual
> > # Config.in.$ARCH files
> > config BR2_ARCH
> >
>
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 2/8] package/gcc: hide versions too old for the current arch
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 1/8] arch: introduce minimal required " Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 3/8] toolchain/external-custom: " Yann E. MORIN
` (7 subsequent siblings)
9 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Begin the conversion from hard-coded dependencies on architectures, to
architecture-specified version requirement, using the newly introduced
BR2_ARCH_NEEDS_GCC_AT_LEAST_XXX symbols.
Hard-coded dependencies will be removed progressively, as archs are
individually coonverted over to using the new symbols.
We do not change the architecture-specific versions for arc and
openrisc, because there is no point in doing so for those, as they
use special, non-upstream versions anyway.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
---
package/gcc/Config.in.host | 3 +++
1 file changed, 3 insertions(+)
diff --git a/package/gcc/Config.in.host b/package/gcc/Config.in.host
index ec7b6924ee..69bebc1f9e 100644
--- a/package/gcc/Config.in.host
+++ b/package/gcc/Config.in.host
@@ -22,6 +22,7 @@ config BR2_GCC_VERSION_OR1K
config BR2_GCC_VERSION_4_9_X
bool "gcc 4.9.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
# Broken or unsupported architectures
depends on !BR2_arc && !BR2_bfin && !BR2_or1k
# Broken or unsupported ARM cores
@@ -43,6 +44,7 @@ config BR2_GCC_VERSION_4_9_X
config BR2_GCC_VERSION_5_X
bool "gcc 5.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
# Broken or unsupported architectures
depends on !BR2_arc && !BR2_bfin && !BR2_or1k
# Broken or unsupported ARM cores
@@ -57,6 +59,7 @@ config BR2_GCC_VERSION_5_X
config BR2_GCC_VERSION_6_X
bool "gcc 6.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
# Broken or unsupported architectures
depends on !BR2_arc
depends on !BR2_or1k
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 3/8] toolchain/external-custom: hide versions too old for the current arch
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 1/8] arch: introduce minimal required " Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 2/8] package/gcc: hide versions too old for the current arch Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-10-07 9:27 ` Romain Naour
2017-09-03 9:44 ` [Buildroot] [PATCH 4/8] toolchain/external: " Yann E. MORIN
` (6 subsequent siblings)
9 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
When an architecture expresses a requirement on the gcc version, limit
the version choice in the custom external toolchain.
The ratioanle being that there is no point in offering that version to
the user if we know before-hand that the gcc version will not work for
that architecture.
All versions below the minimum we support is just made conditional to
that minimum as well, including the "older" entry.
However, this means that the "older" entry is no longer available when
the architecture requires a minimum gcc version. A user who wants to use
a toolchain with a gcc older than the minimum will have no choice but to
realise the toolchain is not suitable (or lie and we would catch that
when checking the gcc version anyway).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
Note: this implies that it is no longer possible to use a custom
external toolchain in which a specific core has been back-ported
Thus, this patch may cause regressions in very specific and rare
corner cases. Do we want to support those cases? If yes, then we
should drop this patch.
---
.../toolchain-external-custom/Config.in.options | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
index ae343ddad5..eb7c30f171 100644
--- a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
+++ b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
@@ -18,6 +18,11 @@ config BR2_TOOLCHAIN_EXTERNAL_CUSTOM_PREFIX
choice
bool "External toolchain gcc version"
+ default BR2_TOOLCHAIN_EXTERNAL_GCC_7 if BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+ default BR2_TOOLCHAIN_EXTERNAL_GCC_6 if BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+ default BR2_TOOLCHAIN_EXTERNAL_GCC_5 if BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+ default BR2_TOOLCHAIN_EXTERNAL_GCC_4_9 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
+ default BR2_TOOLCHAIN_EXTERNAL_GCC_4_8 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
default BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
help
Set to the gcc version that is used by your external
@@ -29,42 +34,52 @@ config BR2_TOOLCHAIN_EXTERNAL_GCC_7
config BR2_TOOLCHAIN_EXTERNAL_GCC_6
bool "6.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
select BR2_TOOLCHAIN_GCC_AT_LEAST_6
config BR2_TOOLCHAIN_EXTERNAL_GCC_5
bool "5.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
select BR2_TOOLCHAIN_GCC_AT_LEAST_5
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_9
bool "4.9.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_8
bool "4.8.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_8
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_7
bool "4.7.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_7
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_6
bool "4.6.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_6
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_5
bool "4.5.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_5
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_4
bool "4.4.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_4
config BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
bool "4.3.x"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_3
config BR2_TOOLCHAIN_EXTERNAL_GCC_OLD
bool "older"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
help
Use this option if your GCC version is older than any of the
above.
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 3/8] toolchain/external-custom: hide versions too old for the current arch
2017-09-03 9:44 ` [Buildroot] [PATCH 3/8] toolchain/external-custom: " Yann E. MORIN
@ 2017-10-07 9:27 ` Romain Naour
2017-10-07 12:27 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Romain Naour @ 2017-10-07 9:27 UTC (permalink / raw)
To: buildroot
Hi Yann, All
Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
> When an architecture expresses a requirement on the gcc version, limit
> the version choice in the custom external toolchain.
>
> The ratioanle being that there is no point in offering that version to
rationale
> the user if we know before-hand that the gcc version will not work for
> that architecture.
>
> All versions below the minimum we support is just made conditional to
> that minimum as well, including the "older" entry.
>
> However, this means that the "older" entry is no longer available when
> the architecture requires a minimum gcc version. A user who wants to use
> a toolchain with a gcc older than the minimum will have no choice but to
> realise the toolchain is not suitable (or lie and we would catch that
realize
> when checking the gcc version anyway).
>
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
>
> ---
> Note: this implies that it is no longer possible to use a custom
> external toolchain in which a specific core has been back-ported
> Thus, this patch may cause regressions in very specific and rare
> corner cases. Do we want to support those cases? If yes, then we
> should drop this patch.
This can be a problem for Linaro toolchains [1] where the GCC support for some
new cpu are backported. (Ok this example is moot since Buildroot doesn't support
these cpu).
Also since BR2_TOOLCHAIN_EXTERNAL_GCC_XXX is an user choice, we can print a
warning instead of completely disabling the gcc version ?
Best regards,
Romain
[1] https://releases.linaro.org/components/toolchain/binaries/5.4-2017.05/
> ---
> .../toolchain-external-custom/Config.in.options | 15 +++++++++++++++
> 1 file changed, 15 insertions(+)
>
> diff --git a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> index ae343ddad5..eb7c30f171 100644
> --- a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> +++ b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> @@ -18,6 +18,11 @@ config BR2_TOOLCHAIN_EXTERNAL_CUSTOM_PREFIX
>
> choice
> bool "External toolchain gcc version"
> + default BR2_TOOLCHAIN_EXTERNAL_GCC_7 if BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> + default BR2_TOOLCHAIN_EXTERNAL_GCC_6 if BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> + default BR2_TOOLCHAIN_EXTERNAL_GCC_5 if BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> + default BR2_TOOLCHAIN_EXTERNAL_GCC_4_9 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> + default BR2_TOOLCHAIN_EXTERNAL_GCC_4_8 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> default BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
> help
> Set to the gcc version that is used by your external
> @@ -29,42 +34,52 @@ config BR2_TOOLCHAIN_EXTERNAL_GCC_7
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_6
> bool "6.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> select BR2_TOOLCHAIN_GCC_AT_LEAST_6
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_5
> bool "5.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> select BR2_TOOLCHAIN_GCC_AT_LEAST_5
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_9
> bool "4.9.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_8
> bool "4.8.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_8
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_7
> bool "4.7.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_7
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_6
> bool "4.6.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_6
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_5
> bool "4.5.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_5
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_4
> bool "4.4.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_4
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
> bool "4.3.x"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_3
>
> config BR2_TOOLCHAIN_EXTERNAL_GCC_OLD
> bool "older"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> help
> Use this option if your GCC version is older than any of the
> above.
>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 3/8] toolchain/external-custom: hide versions too old for the current arch
2017-10-07 9:27 ` Romain Naour
@ 2017-10-07 12:27 ` Yann E. MORIN
0 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-10-07 12:27 UTC (permalink / raw)
To: buildroot
Romain, All,
On 2017-10-07 11:27 +0200, Romain Naour spake thusly:
> Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
[--SNIP--]
> > Note: this implies that it is no longer possible to use a custom
> > external toolchain in which a specific core has been back-ported
> > Thus, this patch may cause regressions in very specific and rare
> > corner cases. Do we want to support those cases? If yes, then we
> > should drop this patch.
>
> This can be a problem for Linaro toolchains [1] where the GCC support for some
> new cpu are backported. (Ok this example is moot since Buildroot doesn't support
> these cpu).
On the other hand, we could get a non-working configuration for other
external toolchains.
And that is not nice to let the user generate a configuration we know if
invalid.
> Also since BR2_TOOLCHAIN_EXTERNAL_GCC_XXX is an user choice, we can print a
> warning instead of completely disabling the gcc version ?
I'll see if I can do something about that.
Regards,
Yann E. MORIN.
> Best regards,
> Romain
>
> [1] https://releases.linaro.org/components/toolchain/binaries/5.4-2017.05/
>
> > ---
> > .../toolchain-external-custom/Config.in.options | 15 +++++++++++++++
> > 1 file changed, 15 insertions(+)
> >
> > diff --git a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> > index ae343ddad5..eb7c30f171 100644
> > --- a/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> > +++ b/toolchain/toolchain-external/toolchain-external-custom/Config.in.options
> > @@ -18,6 +18,11 @@ config BR2_TOOLCHAIN_EXTERNAL_CUSTOM_PREFIX
> >
> > choice
> > bool "External toolchain gcc version"
> > + default BR2_TOOLCHAIN_EXTERNAL_GCC_7 if BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > + default BR2_TOOLCHAIN_EXTERNAL_GCC_6 if BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > + default BR2_TOOLCHAIN_EXTERNAL_GCC_5 if BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > + default BR2_TOOLCHAIN_EXTERNAL_GCC_4_9 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > + default BR2_TOOLCHAIN_EXTERNAL_GCC_4_8 if BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > default BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
> > help
> > Set to the gcc version that is used by your external
> > @@ -29,42 +34,52 @@ config BR2_TOOLCHAIN_EXTERNAL_GCC_7
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_6
> > bool "6.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_6
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_5
> > bool "5.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_5
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_9
> > bool "4.9.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_8
> > bool "4.8.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_8
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_7
> > bool "4.7.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_7
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_6
> > bool "4.6.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_6
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_5
> > bool "4.5.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_5
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_4
> > bool "4.4.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_4
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_4_3
> > bool "4.3.x"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > select BR2_TOOLCHAIN_GCC_AT_LEAST_4_3
> >
> > config BR2_TOOLCHAIN_EXTERNAL_GCC_OLD
> > bool "older"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8
> > help
> > Use this option if your GCC version is older than any of the
> > above.
> >
>
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/8] toolchain/external: hide versions too old for the current arch
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (2 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 3/8] toolchain/external-custom: " Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-10-07 9:57 ` Romain Naour
2017-09-03 9:44 ` [Buildroot] [PATCH 5/8] arch/bfin: needs gcc >= 6 Yann E. MORIN
` (5 subsequent siblings)
9 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Hide the toolchains if the arch requires a gcc version more recent
than the one they provide.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
.../toolchain-external/toolchain-external-codescape-img-mips/Config.in | 1 +
.../toolchain-external/toolchain-external-codescape-mti-mips/Config.in | 1 +
.../toolchain-external-codesourcery-aarch64/Config.in | 1 +
.../toolchain-external/toolchain-external-codesourcery-amd64/Config.in | 1 +
.../toolchain-external/toolchain-external-codesourcery-arm/Config.in | 2 ++
.../toolchain-external/toolchain-external-codesourcery-mips/Config.in | 2 ++
.../toolchain-external/toolchain-external-codesourcery-niosII/Config.in | 1 +
.../toolchain-external/toolchain-external-linaro-aarch64/Config.in | 1 +
toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in | 2 ++
toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in | 2 ++
toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in | 1 +
11 files changed, 15 insertions(+)
diff --git a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
index e29c4dcb87..8b2eab0ad2 100644
--- a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_IMG_MIPS
bool "Codescape IMG GNU Linux Toolchain 2016.05"
depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_MIPS_CPU_MIPS32R6 || (BR2_MIPS_CPU_MIPS64R6 && !BR2_MIPS_SOFT_FLOAT)
select BR2_TOOLCHAIN_EXTERNAL_GLIBC
diff --git a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
index efe6f8527e..db191fd1be 100644
--- a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_MTI_MIPS
bool "Codescape MTI GNU Linux Toolchain 2016.05"
depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_MIPS_CPU_MIPS32R2 || (BR2_MIPS_CPU_MIPS64R2 && !BR2_MIPS_SOFT_FLOAT) || \
BR2_MIPS_CPU_MIPS32R5 || (BR2_MIPS_CPU_MIPS64R5 && !BR2_MIPS_SOFT_FLOAT)
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
index 66a032e9ac..24cbcffa29 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AARCH64
bool "CodeSourcery AArch64 2014.11"
depends on BR2_aarch64
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
# a57/a53 and a72/a53 appeared in gcc-6 or were broken before
depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
index 101e227af9..50c5a20019 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AMD64
bool "Sourcery CodeBench AMD64 2016.11"
depends on BR2_x86_64
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on !BR2_STATIC_LIBS
depends on BR2_x86_jaguar || BR2_x86_steamroller
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
index e1a7891007..b91daa5c81 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
bool "Sourcery CodeBench ARM 2014.05"
depends on BR2_arm
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
# a15/a7 appeared in gcc-4.9, a17/a7 in gcc-5, a57/a53 and a72/a53
# in gcc-6, or they each were broken earlier than that.
depends on !BR2_cortex_a15_a7 && !BR2_cortex_a17_a7
@@ -35,5 +36,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
comment "Sourcery CodeBench toolchains available for the EABI ABI"
depends on BR2_arm
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
depends on !BR2_ARM_EABI
depends on !BR2_STATIC_LIBS
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
index 6a13ae6cd6..734f0e4c7c 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
@@ -2,6 +2,7 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
bool "Sourcery CodeBench MIPS 2016.05"
depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
depends on BR2_MIPS_CPU_MIPS32R2 || BR2_MIPS_CPU_MIPS64R2
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
# Unsupported MIPS cores
depends on !BR2_mips_interaptiv
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
@@ -89,5 +90,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
Select BR2_SOFT_FLOAT
comment "Sourcery CodeBench toolchains are only available for MIPS/MIPS64 o32 and n64"
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
depends on BR2_MIPS_NABI32
depends on !BR2_STATIC_LIBS
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
index 341bc2ab44..7563995eb6 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_NIOSII
bool "Sourcery CodeBench Nios-II 2017.05"
depends on BR2_nios2
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on !BR2_STATIC_LIBS
select BR2_TOOLCHAIN_EXTERNAL_GLIBC
diff --git a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
index caa5ed340a..71ae73adb8 100644
--- a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_LINARO_AARCH64
bool "Linaro AArch64 2017.02"
depends on BR2_aarch64
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on !BR2_STATIC_LIBS
select BR2_TOOLCHAIN_EXTERNAL_GLIBC
diff --git a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
index f538eb7eb3..f5bbbaaad8 100644
--- a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
@@ -1,11 +1,13 @@
comment "Linaro toolchains available for Cortex-A + EABIhf"
depends on BR2_arm
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
depends on !BR2_STATIC_LIBS
config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM
bool "Linaro ARM 2017.02"
depends on BR2_arm
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABIHF
diff --git a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
index 14c7481f81..64e97eeea7 100644
--- a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
@@ -1,11 +1,13 @@
comment "Linaro toolchains available for Cortex-A + EABIhf"
depends on BR2_armeb
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
depends on !BR2_STATIC_LIBS
config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARMEB
bool "Linaro armeb 2017.02"
depends on BR2_armeb
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABIHF
diff --git a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
index f438ea765d..8e4344a83d 100644
--- a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
@@ -1,6 +1,7 @@
config BR2_TOOLCHAIN_EXTERNAL_SYNOPSYS_ARC
bool "Synopsys ARC 2016.09 toolchain"
depends on BR2_arc
+ depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
depends on BR2_HOSTARCH = "x86_64"
select BR2_TOOLCHAIN_EXTERNAL_UCLIBC
select BR2_INSTALL_LIBSTDCPP
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 4/8] toolchain/external: hide versions too old for the current arch
2017-09-03 9:44 ` [Buildroot] [PATCH 4/8] toolchain/external: " Yann E. MORIN
@ 2017-10-07 9:57 ` Romain Naour
2017-10-07 12:15 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Romain Naour @ 2017-10-07 9:57 UTC (permalink / raw)
To: buildroot
Yann, All,
This patch needs to be rebased on master after the last linaro toolchain version
bump (sorry ;-) )
Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
> Hide the toolchains if the arch requires a gcc version more recent
> than the one they provide.
>
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> ---
> .../toolchain-external/toolchain-external-codescape-img-mips/Config.in | 1 +
> .../toolchain-external/toolchain-external-codescape-mti-mips/Config.in | 1 +
> .../toolchain-external-codesourcery-aarch64/Config.in | 1 +
> .../toolchain-external/toolchain-external-codesourcery-amd64/Config.in | 1 +
> .../toolchain-external/toolchain-external-codesourcery-arm/Config.in | 2 ++
> .../toolchain-external/toolchain-external-codesourcery-mips/Config.in | 2 ++
> .../toolchain-external/toolchain-external-codesourcery-niosII/Config.in | 1 +
> .../toolchain-external/toolchain-external-linaro-aarch64/Config.in | 1 +
> toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in | 2 ++
> toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in | 2 ++
> toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in | 1 +
> 11 files changed, 15 insertions(+)
>
> diff --git a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> index e29c4dcb87..8b2eab0ad2 100644
> --- a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_IMG_MIPS
> bool "Codescape IMG GNU Linux Toolchain 2016.05"
> depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
Having
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
and
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
looks redundant...
Also where BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9 is selected when you use a packaged
toolchain ?
With the complete series applied I can't select a packaged external toolchain.
Try with to select the Linaro toolchain with this defconfig:
BR2_arm=y
BR2_cortex_a9=y
BR2_ARM_ENABLE_NEON=y
BR2_ARM_ENABLE_VFP=y
BR2_TOOLCHAIN_EXTERNAL=y
Best regards,
Romain
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on BR2_MIPS_CPU_MIPS32R6 || (BR2_MIPS_CPU_MIPS64R6 && !BR2_MIPS_SOFT_FLOAT)
> select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> diff --git a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> index efe6f8527e..db191fd1be 100644
> --- a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_MTI_MIPS
> bool "Codescape MTI GNU Linux Toolchain 2016.05"
> depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on BR2_MIPS_CPU_MIPS32R2 || (BR2_MIPS_CPU_MIPS64R2 && !BR2_MIPS_SOFT_FLOAT) || \
> BR2_MIPS_CPU_MIPS32R5 || (BR2_MIPS_CPU_MIPS64R5 && !BR2_MIPS_SOFT_FLOAT)
> diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> index 66a032e9ac..24cbcffa29 100644
> --- a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AARCH64
> bool "CodeSourcery AArch64 2014.11"
> depends on BR2_aarch64
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> # a57/a53 and a72/a53 appeared in gcc-6 or were broken before
> depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> index 101e227af9..50c5a20019 100644
> --- a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AMD64
> bool "Sourcery CodeBench AMD64 2016.11"
> depends on BR2_x86_64
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on !BR2_STATIC_LIBS
> depends on BR2_x86_jaguar || BR2_x86_steamroller
> diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> index e1a7891007..b91daa5c81 100644
> --- a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
> bool "Sourcery CodeBench ARM 2014.05"
> depends on BR2_arm
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> # a15/a7 appeared in gcc-4.9, a17/a7 in gcc-5, a57/a53 and a72/a53
> # in gcc-6, or they each were broken earlier than that.
> depends on !BR2_cortex_a15_a7 && !BR2_cortex_a17_a7
> @@ -35,5 +36,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
>
> comment "Sourcery CodeBench toolchains available for the EABI ABI"
> depends on BR2_arm
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> depends on !BR2_ARM_EABI
> depends on !BR2_STATIC_LIBS
> diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> index 6a13ae6cd6..734f0e4c7c 100644
> --- a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> @@ -2,6 +2,7 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
> bool "Sourcery CodeBench MIPS 2016.05"
> depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
> depends on BR2_MIPS_CPU_MIPS32R2 || BR2_MIPS_CPU_MIPS64R2
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> # Unsupported MIPS cores
> depends on !BR2_mips_interaptiv
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> @@ -89,5 +90,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
> Select BR2_SOFT_FLOAT
>
> comment "Sourcery CodeBench toolchains are only available for MIPS/MIPS64 o32 and n64"
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> depends on BR2_MIPS_NABI32
> depends on !BR2_STATIC_LIBS
> diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> index 341bc2ab44..7563995eb6 100644
> --- a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_NIOSII
> bool "Sourcery CodeBench Nios-II 2017.05"
> depends on BR2_nios2
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on !BR2_STATIC_LIBS
> select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> diff --git a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> index caa5ed340a..71ae73adb8 100644
> --- a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_LINARO_AARCH64
> bool "Linaro AArch64 2017.02"
> depends on BR2_aarch64
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on !BR2_STATIC_LIBS
> select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> diff --git a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> index f538eb7eb3..f5bbbaaad8 100644
> --- a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> @@ -1,11 +1,13 @@
> comment "Linaro toolchains available for Cortex-A + EABIhf"
> depends on BR2_arm
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
> depends on !BR2_STATIC_LIBS
>
> config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM
> bool "Linaro ARM 2017.02"
> depends on BR2_arm
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on BR2_ARM_EABIHF
> diff --git a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> index 14c7481f81..64e97eeea7 100644
> --- a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> @@ -1,11 +1,13 @@
> comment "Linaro toolchains available for Cortex-A + EABIhf"
> depends on BR2_armeb
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
> depends on !BR2_STATIC_LIBS
>
> config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARMEB
> bool "Linaro armeb 2017.02"
> depends on BR2_armeb
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
> depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> depends on BR2_ARM_EABIHF
> diff --git a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> index f438ea765d..8e4344a83d 100644
> --- a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> +++ b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> @@ -1,6 +1,7 @@
> config BR2_TOOLCHAIN_EXTERNAL_SYNOPSYS_ARC
> bool "Synopsys ARC 2016.09 toolchain"
> depends on BR2_arc
> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> depends on BR2_HOSTARCH = "x86_64"
> select BR2_TOOLCHAIN_EXTERNAL_UCLIBC
> select BR2_INSTALL_LIBSTDCPP
>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/8] toolchain/external: hide versions too old for the current arch
2017-10-07 9:57 ` Romain Naour
@ 2017-10-07 12:15 ` Yann E. MORIN
2017-10-07 19:55 ` Romain Naour
0 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-10-07 12:15 UTC (permalink / raw)
To: buildroot
Romain, All,
On 2017-10-07 11:57 +0200, Romain Naour spake thusly:
> This patch needs to be rebased on master after the last linaro toolchain version
> bump (sorry ;-) )
>
> Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
> > Hide the toolchains if the arch requires a gcc version more recent
> > than the one they provide.
> >
> > Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> > Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> > ---
> > .../toolchain-external/toolchain-external-codescape-img-mips/Config.in | 1 +
> > .../toolchain-external/toolchain-external-codescape-mti-mips/Config.in | 1 +
> > .../toolchain-external-codesourcery-aarch64/Config.in | 1 +
> > .../toolchain-external/toolchain-external-codesourcery-amd64/Config.in | 1 +
> > .../toolchain-external/toolchain-external-codesourcery-arm/Config.in | 2 ++
> > .../toolchain-external/toolchain-external-codesourcery-mips/Config.in | 2 ++
> > .../toolchain-external/toolchain-external-codesourcery-niosII/Config.in | 1 +
> > .../toolchain-external/toolchain-external-linaro-aarch64/Config.in | 1 +
> > toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in | 2 ++
> > toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in | 2 ++
> > toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in | 1 +
> > 11 files changed, 15 insertions(+)
> >
> > diff --git a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> > index e29c4dcb87..8b2eab0ad2 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_IMG_MIPS
> > bool "Codescape IMG GNU Linux Toolchain 2016.05"
> > depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
>
> Having
>
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
>
> and
> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
>
> looks redundant...
It might, but this serves two different purposes.
> Also where BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9 is selected when you use a packaged
> toolchain ?
The architectures do. Either the full arch, or individual cores.
> With the complete series applied I can't select a packaged external toolchain.
>
> Try with to select the Linaro toolchain with this defconfig:
>
> BR2_arm=y
> BR2_cortex_a9=y
> BR2_ARM_ENABLE_NEON=y
> BR2_ARM_ENABLE_VFP=y
> BR2_TOOLCHAIN_EXTERNAL=y
Weird, it works for me:
$ make defconfig BR2_DEFCONFIG=$(pwd)/kubu.defconfig
$ grep LINARO .config
BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM=y
Regards,
Yann E. MORIN.
> Best regards,
> Romain
>
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on BR2_MIPS_CPU_MIPS32R6 || (BR2_MIPS_CPU_MIPS64R6 && !BR2_MIPS_SOFT_FLOAT)
> > select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> > diff --git a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> > index efe6f8527e..db191fd1be 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codescape-mti-mips/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_MTI_MIPS
> > bool "Codescape MTI GNU Linux Toolchain 2016.05"
> > depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on BR2_MIPS_CPU_MIPS32R2 || (BR2_MIPS_CPU_MIPS64R2 && !BR2_MIPS_SOFT_FLOAT) || \
> > BR2_MIPS_CPU_MIPS32R5 || (BR2_MIPS_CPU_MIPS64R5 && !BR2_MIPS_SOFT_FLOAT)
> > diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> > index 66a032e9ac..24cbcffa29 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-aarch64/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AARCH64
> > bool "CodeSourcery AArch64 2014.11"
> > depends on BR2_aarch64
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
> > # a57/a53 and a72/a53 appeared in gcc-6 or were broken before
> > depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> > index 101e227af9..50c5a20019 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-amd64/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_AMD64
> > bool "Sourcery CodeBench AMD64 2016.11"
> > depends on BR2_x86_64
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on !BR2_STATIC_LIBS
> > depends on BR2_x86_jaguar || BR2_x86_steamroller
> > diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> > index e1a7891007..b91daa5c81 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
> > bool "Sourcery CodeBench ARM 2014.05"
> > depends on BR2_arm
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > # a15/a7 appeared in gcc-4.9, a17/a7 in gcc-5, a57/a53 and a72/a53
> > # in gcc-6, or they each were broken earlier than that.
> > depends on !BR2_cortex_a15_a7 && !BR2_cortex_a17_a7
> > @@ -35,5 +36,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
> >
> > comment "Sourcery CodeBench toolchains available for the EABI ABI"
> > depends on BR2_arm
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
> > depends on !BR2_ARM_EABI
> > depends on !BR2_STATIC_LIBS
> > diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> > index 6a13ae6cd6..734f0e4c7c 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-mips/Config.in
> > @@ -2,6 +2,7 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
> > bool "Sourcery CodeBench MIPS 2016.05"
> > depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
> > depends on BR2_MIPS_CPU_MIPS32R2 || BR2_MIPS_CPU_MIPS64R2
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > # Unsupported MIPS cores
> > depends on !BR2_mips_interaptiv
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > @@ -89,5 +90,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_MIPS
> > Select BR2_SOFT_FLOAT
> >
> > comment "Sourcery CodeBench toolchains are only available for MIPS/MIPS64 o32 and n64"
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
> > depends on BR2_MIPS_NABI32
> > depends on !BR2_STATIC_LIBS
> > diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> > index 341bc2ab44..7563995eb6 100644
> > --- a/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-codesourcery-niosII/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_NIOSII
> > bool "Sourcery CodeBench Nios-II 2017.05"
> > depends on BR2_nios2
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on !BR2_STATIC_LIBS
> > select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> > diff --git a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> > index caa5ed340a..71ae73adb8 100644
> > --- a/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-linaro-aarch64/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_LINARO_AARCH64
> > bool "Linaro AArch64 2017.02"
> > depends on BR2_aarch64
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on !BR2_STATIC_LIBS
> > select BR2_TOOLCHAIN_EXTERNAL_GLIBC
> > diff --git a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> > index f538eb7eb3..f5bbbaaad8 100644
> > --- a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
> > @@ -1,11 +1,13 @@
> > comment "Linaro toolchains available for Cortex-A + EABIhf"
> > depends on BR2_arm
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
> > depends on !BR2_STATIC_LIBS
> >
> > config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM
> > bool "Linaro ARM 2017.02"
> > depends on BR2_arm
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on BR2_ARM_EABIHF
> > diff --git a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> > index 14c7481f81..64e97eeea7 100644
> > --- a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
> > @@ -1,11 +1,13 @@
> > comment "Linaro toolchains available for Cortex-A + EABIhf"
> > depends on BR2_armeb
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on !BR2_ARM_CPU_ARMV7A || !BR2_ARM_EABIHF
> > depends on !BR2_STATIC_LIBS
> >
> > config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARMEB
> > bool "Linaro armeb 2017.02"
> > depends on BR2_armeb
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
> > depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
> > depends on BR2_ARM_EABIHF
> > diff --git a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> > index f438ea765d..8e4344a83d 100644
> > --- a/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> > +++ b/toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in
> > @@ -1,6 +1,7 @@
> > config BR2_TOOLCHAIN_EXTERNAL_SYNOPSYS_ARC
> > bool "Synopsys ARC 2016.09 toolchain"
> > depends on BR2_arc
> > + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
> > depends on BR2_HOSTARCH = "x86_64"
> > select BR2_TOOLCHAIN_EXTERNAL_UCLIBC
> > select BR2_INSTALL_LIBSTDCPP
> >
>
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 4/8] toolchain/external: hide versions too old for the current arch
2017-10-07 12:15 ` Yann E. MORIN
@ 2017-10-07 19:55 ` Romain Naour
0 siblings, 0 replies; 51+ messages in thread
From: Romain Naour @ 2017-10-07 19:55 UTC (permalink / raw)
To: buildroot
Hi Yann, all,
Le 07/10/2017 ? 14:15, Yann E. MORIN a ?crit?:
> Romain, All,
>
> On 2017-10-07 11:57 +0200, Romain Naour spake thusly:
>> This patch needs to be rebased on master after the last linaro toolchain version
>> bump (sorry ;-) )
>>
>> Le 03/09/2017 ? 11:44, Yann E. MORIN a ?crit?:
>>> Hide the toolchains if the arch requires a gcc version more recent
>>> than the one they provide.
>>>
>>> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
>>> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
>>> ---
>>> .../toolchain-external/toolchain-external-codescape-img-mips/Config.in | 1 +
>>> .../toolchain-external/toolchain-external-codescape-mti-mips/Config.in | 1 +
>>> .../toolchain-external-codesourcery-aarch64/Config.in | 1 +
>>> .../toolchain-external/toolchain-external-codesourcery-amd64/Config.in | 1 +
>>> .../toolchain-external/toolchain-external-codesourcery-arm/Config.in | 2 ++
>>> .../toolchain-external/toolchain-external-codesourcery-mips/Config.in | 2 ++
>>> .../toolchain-external/toolchain-external-codesourcery-niosII/Config.in | 1 +
>>> .../toolchain-external/toolchain-external-linaro-aarch64/Config.in | 1 +
>>> toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in | 2 ++
>>> toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in | 2 ++
>>> toolchain/toolchain-external/toolchain-external-synopsys-arc/Config.in | 1 +
>>> 11 files changed, 15 insertions(+)
>>>
>>> diff --git a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
>>> index e29c4dcb87..8b2eab0ad2 100644
>>> --- a/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
>>> +++ b/toolchain/toolchain-external/toolchain-external-codescape-img-mips/Config.in
>>> @@ -1,6 +1,7 @@
>>> config BR2_TOOLCHAIN_EXTERNAL_CODESCAPE_IMG_MIPS
>>> bool "Codescape IMG GNU Linux Toolchain 2016.05"
>>> depends on BR2_mips || BR2_mipsel || BR2_mips64 || BR2_mips64el
>>
>> Having
>>
>>> + depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
>>
>> and
>> select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
>>
>> looks redundant...
>
> It might, but this serves two different purposes.
Yes, I understand that.
>
>> Also where BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9 is selected when you use a packaged
>> toolchain ?
>
> The architectures do. Either the full arch, or individual cores.
>
>> With the complete series applied I can't select a packaged external toolchain.
>>
>> Try with to select the Linaro toolchain with this defconfig:
>>
>> BR2_arm=y
>> BR2_cortex_a9=y
>> BR2_ARM_ENABLE_NEON=y
>> BR2_ARM_ENABLE_VFP=y
>> BR2_TOOLCHAIN_EXTERNAL=y
>
> Weird, it works for me:
>
> $ make defconfig BR2_DEFCONFIG=$(pwd)/kubu.defconfig
> $ grep LINARO .config
> BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM=y
I made a mistake while applying the patch locally and resolve the conflict due
to the Linaro toolchain bump.
The '!' was missing before "BR2_ARCH_NEEDS_GCC_AT_LEAST_7"...
That's why I wasn't able to select the toolchain.
Sorry for the noise.
Best regards,
Romain
>
> Regards,
> Yann E. MORIN.
>
>> Best regards,
>> Romain
>>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 5/8] arch/bfin: needs gcc >= 6
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (3 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 4/8] toolchain/external: " Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 6/8] arch/mips: some variants need different gcc versions Yann E. MORIN
` (4 subsequent siblings)
9 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in | 1 +
package/gcc/Config.in.host | 4 ++--
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/arch/Config.in b/arch/Config.in
index 607da36c1d..d3f63da34f 100644
--- a/arch/Config.in
+++ b/arch/Config.in
@@ -79,6 +79,7 @@ config BR2_aarch64_be
config BR2_bfin
bool "Blackfin"
select BR2_ARCH_HAS_FDPIC_SUPPORT
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
help
The Blackfin is a family of 16 or 32-bit microprocessors developed,
manufactured and marketed by Analog Devices.
diff --git a/package/gcc/Config.in.host b/package/gcc/Config.in.host
index 69bebc1f9e..df836d96e4 100644
--- a/package/gcc/Config.in.host
+++ b/package/gcc/Config.in.host
@@ -24,7 +24,7 @@ config BR2_GCC_VERSION_4_9_X
bool "gcc 4.9.x"
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
# Broken or unsupported architectures
- depends on !BR2_arc && !BR2_bfin && !BR2_or1k
+ depends on !BR2_arc && !BR2_or1k
# Broken or unsupported ARM cores
depends on !BR2_cortex_a17 && !BR2_cortex_a17_a7
depends on !BR2_cortex_a72 && !BR2_cortex_a72_a53
@@ -46,7 +46,7 @@ config BR2_GCC_VERSION_5_X
bool "gcc 5.x"
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
# Broken or unsupported architectures
- depends on !BR2_arc && !BR2_bfin && !BR2_or1k
+ depends on !BR2_arc && !BR2_or1k
# Broken or unsupported ARM cores
depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
# musl ppc64 unsupported
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 6/8] arch/mips: some variants need different gcc versions
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (4 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 5/8] arch/bfin: needs gcc >= 6 Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 7/8] arch/arm: " Yann E. MORIN
` (3 subsequent siblings)
9 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
We use the conditions currently expressed in the gcc version choice.
We leave the msul vs mips64 conditions in gcc, because the "fault"
really is on gcc, which does not recognise the mips64+musl tuples,
so the fix lie with gcc, and the current conditions are fitting.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.mips | 7 +++++++
package/gcc/Config.in.host | 8 --------
2 files changed, 7 insertions(+), 8 deletions(-)
diff --git a/arch/Config.in.mips b/arch/Config.in.mips
index 1cce1710da..c08b2a2968 100644
--- a/arch/Config.in.mips
+++ b/arch/Config.in.mips
@@ -7,9 +7,11 @@ config BR2_MIPS_CPU_MIPS32R2
select BR2_MIPS_NAN_LEGACY
config BR2_MIPS_CPU_MIPS32R5
bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_MIPS_CPU_MIPS32R6
bool
select BR2_MIPS_NAN_2008
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_MIPS_CPU_MIPS64
bool
select BR2_MIPS_NAN_LEGACY
@@ -18,9 +20,11 @@ config BR2_MIPS_CPU_MIPS64R2
select BR2_MIPS_NAN_LEGACY
config BR2_MIPS_CPU_MIPS64R5
bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_MIPS_CPU_MIPS64R6
bool
select BR2_MIPS_NAN_2008
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
choice
prompt "Target Architecture Variant"
@@ -53,11 +57,13 @@ config BR2_mips_interaptiv
bool "interAptiv"
depends on !BR2_ARCH_IS_64
select BR2_MIPS_CPU_MIPS32R2
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_mips_m5150
bool "M5150"
depends on !BR2_ARCH_IS_64
select BR2_MIPS_CPU_MIPS32R5
select BR2_MIPS_NAN_2008
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_mips_m6250
bool "M6250"
depends on !BR2_ARCH_IS_64
@@ -101,6 +107,7 @@ config BR2_mips_i6400
bool "I6400"
depends on BR2_ARCH_IS_64
select BR2_MIPS_CPU_MIPS64R6
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_mips_p6600
bool "P6600"
depends on BR2_ARCH_IS_64
diff --git a/package/gcc/Config.in.host b/package/gcc/Config.in.host
index df836d96e4..8c07c29c2d 100644
--- a/package/gcc/Config.in.host
+++ b/package/gcc/Config.in.host
@@ -28,12 +28,6 @@ config BR2_GCC_VERSION_4_9_X
# Broken or unsupported ARM cores
depends on !BR2_cortex_a17 && !BR2_cortex_a17_a7
depends on !BR2_cortex_a72 && !BR2_cortex_a72_a53
- # Unsupported MIPS cores
- depends on !BR2_mips_interaptiv
- # Unsupported for MIPS R5
- depends on !BR2_MIPS_CPU_MIPS32R5 && !BR2_MIPS_CPU_MIPS64R5
- # Unsupported for MIPS R6
- depends on !BR2_MIPS_CPU_MIPS32R6 && !BR2_MIPS_CPU_MIPS64R6
# musl microblaze unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_microblazeel || BR2_microblazebe))
# musl ppc64 unsupported
@@ -51,8 +45,6 @@ config BR2_GCC_VERSION_5_X
depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
# musl ppc64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_powerpc64 || BR2_powerpc64le))
- # Unsupported MIPS cores
- depends on !BR2_mips_interaptiv && !BR2_mips_m5150 && !BR2_mips_i6400
# musl mips64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_mips64 || BR2_mips64el))
select BR2_TOOLCHAIN_GCC_AT_LEAST_5
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 7/8] arch/arm: some variants need different gcc versions
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (5 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 6/8] arch/mips: some variants need different gcc versions Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-09-03 9:44 ` [Buildroot] [PATCH 8/8] package/gcc: slight cleanup and reorg in remaining arch depends Yann E. MORIN
` (2 subsequent siblings)
9 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Take the conditions currently specified in the gcc version choice.
Also, the conditions explained in the commit log for 78c2a9f7 were not
all properly applied, especially the a57-a53 combo needs gcc-6, but
78c2a9f7 forgot to add the condition to gcc-4.9.
gcc-4.9 was excluded for cortex-a17 and a72, but the CodeSourcery
external toolchain, which uses 4.8, was not excluded for those two
cores. Now it is.
Remove the arch condition from gcc and the external toolchains.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 6 ++++++
package/gcc/Config.in.host | 5 -----
.../toolchain-external-codesourcery-arm/Config.in | 4 ----
3 files changed, 6 insertions(+), 9 deletions(-)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 09916df7ad..b58be6bce0 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -190,6 +190,7 @@ config BR2_cortex_a15_a7
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
depends on !BR2_ARCH_IS_64
config BR2_cortex_a17
bool "cortex-A17"
@@ -199,6 +200,7 @@ config BR2_cortex_a17
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
depends on !BR2_ARCH_IS_64
config BR2_cortex_a17_a7
bool "cortex-A17/A7 big.LITTLE"
@@ -208,6 +210,7 @@ config BR2_cortex_a17_a7
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
depends on !BR2_ARCH_IS_64
config BR2_cortex_a53
bool "cortex-A53"
@@ -233,6 +236,7 @@ config BR2_cortex_a57_a53
select BR2_ARM_CPU_HAS_FP_ARMV8
select BR2_ARM_CPU_ARMV8
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_cortex_a72
bool "cortex-A72"
select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
@@ -241,6 +245,7 @@ config BR2_cortex_a72
select BR2_ARM_CPU_HAS_FP_ARMV8
select BR2_ARM_CPU_ARMV8
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_cortex_a72_a53
bool "cortex-A72/A53 big.LITTLE"
select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
@@ -249,6 +254,7 @@ config BR2_cortex_a72_a53
select BR2_ARM_CPU_HAS_FP_ARMV8
select BR2_ARM_CPU_ARMV8
select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_cortex_m3
bool "cortex-M3"
select BR2_ARM_CPU_HAS_THUMB2
diff --git a/package/gcc/Config.in.host b/package/gcc/Config.in.host
index 8c07c29c2d..d8479008d8 100644
--- a/package/gcc/Config.in.host
+++ b/package/gcc/Config.in.host
@@ -25,9 +25,6 @@ config BR2_GCC_VERSION_4_9_X
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
# Broken or unsupported architectures
depends on !BR2_arc && !BR2_or1k
- # Broken or unsupported ARM cores
- depends on !BR2_cortex_a17 && !BR2_cortex_a17_a7
- depends on !BR2_cortex_a72 && !BR2_cortex_a72_a53
# musl microblaze unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_microblazeel || BR2_microblazebe))
# musl ppc64 unsupported
@@ -41,8 +38,6 @@ config BR2_GCC_VERSION_5_X
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
# Broken or unsupported architectures
depends on !BR2_arc && !BR2_or1k
- # Broken or unsupported ARM cores
- depends on !BR2_cortex_a57_a53 && !BR2_cortex_a72_a53
# musl ppc64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_powerpc64 || BR2_powerpc64le))
# musl mips64 unsupported
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
index b91daa5c81..60d86c33d5 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
@@ -2,10 +2,6 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
bool "Sourcery CodeBench ARM 2014.05"
depends on BR2_arm
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
- # a15/a7 appeared in gcc-4.9, a17/a7 in gcc-5, a57/a53 and a72/a53
- # in gcc-6, or they each were broken earlier than that.
- depends on !BR2_cortex_a15_a7 && !BR2_cortex_a17_a7
- depends on !BR2_cortex_a57_53 && !BR2_cortex_a72_53
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABI
# Unsupported ARM cores
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 8/8] package/gcc: slight cleanup and reorg in remaining arch depends
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (6 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 7/8] arch/arm: " Yann E. MORIN
@ 2017-09-03 9:44 ` Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
2017-11-24 21:23 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Thomas Petazzoni
9 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:44 UTC (permalink / raw)
To: buildroot
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
package/gcc/Config.in.host | 13 ++++++-------
1 file changed, 6 insertions(+), 7 deletions(-)
diff --git a/package/gcc/Config.in.host b/package/gcc/Config.in.host
index d8479008d8..4ceb5b18c7 100644
--- a/package/gcc/Config.in.host
+++ b/package/gcc/Config.in.host
@@ -24,12 +24,11 @@ config BR2_GCC_VERSION_4_9_X
bool "gcc 4.9.x"
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_5
# Broken or unsupported architectures
- depends on !BR2_arc && !BR2_or1k
- # musl microblaze unsupported
+ depends on !BR2_arc
+ depends on !BR2_or1k
+ # musl on microblaze, ppc64 and mips64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_microblazeel || BR2_microblazebe))
- # musl ppc64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_powerpc64 || BR2_powerpc64le))
- # musl mips64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_mips64 || BR2_mips64el))
select BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
@@ -37,10 +36,10 @@ config BR2_GCC_VERSION_5_X
bool "gcc 5.x"
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_6
# Broken or unsupported architectures
- depends on !BR2_arc && !BR2_or1k
- # musl ppc64 unsupported
+ depends on !BR2_arc
+ depends on !BR2_or1k
+ # musl on ppc64 and mips64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_powerpc64 || BR2_powerpc64le))
- # musl mips64 unsupported
depends on !(BR2_TOOLCHAIN_USES_MUSL && (BR2_mips64 || BR2_mips64el))
select BR2_TOOLCHAIN_GCC_AT_LEAST_5
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (7 preceding siblings ...)
2017-09-03 9:44 ` [Buildroot] [PATCH 8/8] package/gcc: slight cleanup and reorg in remaining arch depends Yann E. MORIN
@ 2017-09-03 9:53 ` Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic Yann E. MORIN
` (4 more replies)
2017-11-24 21:23 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Thomas Petazzoni
9 siblings, 5 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:53 UTC (permalink / raw)
To: buildroot
Hello All!
Not all architectures we support have support in upstream gcc, glibc,
binutils, and/or linux. In some cases, only a few specific cores do not
have that support.
Furthermore, some architecture, specific cores, or core behaviour/setup
(like the MIPS NaN support) got support in a specific gcc version.
Currently, we have to account for this in at least four locations:
- the architecture
- the toolchain backend choice (internal or externsal)
- the gcc version choice
- each external toolchain
This third series fixes the way we handle NaN and floating point in the
MIPS case.
Currently, they both depend on the gcc version. However, this is
opposite what we usually do in Buildroot: we let the user select their
hardware settings, and then limit the software list according to that
which suports the user's hardware.
So, now that we have a mean for architectures to specify the minimum gcc
version they require, we use that to limit the gcc version depending on
the NaN and FP settings the user did set.
And this concludes this three-series series (a meta-series, made up of
three series. It's series all the way down! Muahaha! ;-) )
Regards,
Yann E. MORIN.
The following changes since commit 41eeac74f986790fe26f03ebf05322120d2c04da
package/gcc: slight cleanup and reorg in remaining arch depends (2017-09-03 11:14:53 +0200)
are available in the git repository at:
git://git.buildroot.org/~ymorin/git/buildroot.git
for you to fetch changes up to a191b26508b5cddafe1059f13fcd2aa4b2ebd689
toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008 (2017-09-03 11:49:02 +0200)
----------------------------------------------------------------
Yann E. MORIN (3):
arch/mips: inverse the NaN logic
arch/mips: inverse the mfpxx logic
toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008
arch/Config.in.mips | 6 +++---
package/gcc/gcc.mk | 2 ++
toolchain/toolchain-buildroot/Config.in | 1 +
toolchain/toolchain-common.in | 4 ----
4 files changed, 6 insertions(+), 7 deletions(-)
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
@ 2017-09-03 9:53 ` Yann E. MORIN
2017-10-07 10:18 ` Romain Naour
2017-09-03 9:53 ` [Buildroot] [PATCH 2/3] arch/mips: inverse the mfpxx logic Yann E. MORIN
` (3 subsequent siblings)
4 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:53 UTC (permalink / raw)
To: buildroot
Currentlym the possibility to choose the NaN encoding is conditional to
having a sufficiently recent gcc version.
Which means that the architecture selection depends on the gcc version.
But that's opposite to what we've always done in Buildroot: the software
versions are conditional to the architecture options. There is nothing
we can do about the hardware: it is there, we can't change it, while we
can restrict ourselves to using software that is working on said
hardware.
Thus, we inverse the logic, to move the condition onto the software
side: whenever NaN-2008 are selected, we restrict the toolchain
selection to at least a gcc-4.9.
But now, the option with the NaN type is always set, so we must enclose
the code in gcc.mk inside a HAS_NAN_OPTION condition, as is already done
for the external toolchain case.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
---
arch/Config.in.mips | 2 +-
package/gcc/gcc.mk | 2 ++
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/arch/Config.in.mips b/arch/Config.in.mips
index c08b2a2968..13d1a477e6 100644
--- a/arch/Config.in.mips
+++ b/arch/Config.in.mips
@@ -176,10 +176,10 @@ config BR2_MIPS_NAN_LEGACY
config BR2_MIPS_NAN_2008
bool
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
choice
prompt "Target NaN"
- depends on BR2_TOOLCHAIN_HAS_MNAN_OPTION
depends on BR2_mips_32r5 || BR2_mips_64r5
default BR2_MIPS_ENABLE_NAN_2008
help
diff --git a/package/gcc/gcc.mk b/package/gcc/gcc.mk
index 49ccccf56f..d45b6d7a06 100644
--- a/package/gcc/gcc.mk
+++ b/package/gcc/gcc.mk
@@ -207,9 +207,11 @@ endif
ifneq ($(call qstrip,$(BR2_GCC_TARGET_ABI)),)
HOST_GCC_COMMON_CONF_OPTS += --with-abi=$(BR2_GCC_TARGET_ABI)
endif
+ifeq ($(BR2_TOOLCHAIN_HAS_MNAN_OPTION),y)
ifneq ($(call qstrip,$(BR2_GCC_TARGET_NAN)),)
HOST_GCC_COMMON_CONF_OPTS += --with-nan=$(BR2_GCC_TARGET_NAN)
endif
+endif
ifneq ($(call qstrip,$(BR2_GCC_TARGET_FP32_MODE)),)
HOST_GCC_COMMON_CONF_OPTS += --with-fp-32=$(BR2_GCC_TARGET_FP32_MODE)
endif
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic
2017-09-03 9:53 ` [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic Yann E. MORIN
@ 2017-10-07 10:18 ` Romain Naour
2017-10-07 12:22 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Romain Naour @ 2017-10-07 10:18 UTC (permalink / raw)
To: buildroot
Yann, All,
Le 03/09/2017 ? 11:53, Yann E. MORIN a ?crit?:
> Currentlym the possibility to choose the NaN encoding is conditional to
> having a sufficiently recent gcc version.
>
> Which means that the architecture selection depends on the gcc version.
>
> But that's opposite to what we've always done in Buildroot: the software
> versions are conditional to the architecture options. There is nothing
> we can do about the hardware: it is there, we can't change it, while we
> can restrict ourselves to using software that is working on said
> hardware.
>
> Thus, we inverse the logic, to move the condition onto the software
> side: whenever NaN-2008 are selected, we restrict the toolchain
> selection to at least a gcc-4.9.
>
> But now, the option with the NaN type is always set, so we must enclose
> the code in gcc.mk inside a HAS_NAN_OPTION condition, as is already done
> for the external toolchain case.
>
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
> ---
> arch/Config.in.mips | 2 +-
> package/gcc/gcc.mk | 2 ++
> 2 files changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/arch/Config.in.mips b/arch/Config.in.mips
> index c08b2a2968..13d1a477e6 100644
> --- a/arch/Config.in.mips
> +++ b/arch/Config.in.mips
> @@ -176,10 +176,10 @@ config BR2_MIPS_NAN_LEGACY
>
> config BR2_MIPS_NAN_2008
> bool
> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
This series require the previous series "arch: some require a minimal gcc
version" adding this option.
Best regards,
Romain
>
> choice
> prompt "Target NaN"
> - depends on BR2_TOOLCHAIN_HAS_MNAN_OPTION
> depends on BR2_mips_32r5 || BR2_mips_64r5
> default BR2_MIPS_ENABLE_NAN_2008
> help
> diff --git a/package/gcc/gcc.mk b/package/gcc/gcc.mk
> index 49ccccf56f..d45b6d7a06 100644
> --- a/package/gcc/gcc.mk
> +++ b/package/gcc/gcc.mk
> @@ -207,9 +207,11 @@ endif
> ifneq ($(call qstrip,$(BR2_GCC_TARGET_ABI)),)
> HOST_GCC_COMMON_CONF_OPTS += --with-abi=$(BR2_GCC_TARGET_ABI)
> endif
> +ifeq ($(BR2_TOOLCHAIN_HAS_MNAN_OPTION),y)
> ifneq ($(call qstrip,$(BR2_GCC_TARGET_NAN)),)
> HOST_GCC_COMMON_CONF_OPTS += --with-nan=$(BR2_GCC_TARGET_NAN)
> endif
> +endif
> ifneq ($(call qstrip,$(BR2_GCC_TARGET_FP32_MODE)),)
> HOST_GCC_COMMON_CONF_OPTS += --with-fp-32=$(BR2_GCC_TARGET_FP32_MODE)
> endif
>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic
2017-10-07 10:18 ` Romain Naour
@ 2017-10-07 12:22 ` Yann E. MORIN
2017-10-07 18:43 ` Romain Naour
0 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-10-07 12:22 UTC (permalink / raw)
To: buildroot
Romain, All,
On 2017-10-07 12:18 +0200, Romain Naour spake thusly:
> Le 03/09/2017 ? 11:53, Yann E. MORIN a ?crit?:
> > Currentlym the possibility to choose the NaN encoding is conditional to
> > having a sufficiently recent gcc version.
> >
> > Which means that the architecture selection depends on the gcc version.
> >
> > But that's opposite to what we've always done in Buildroot: the software
> > versions are conditional to the architecture options. There is nothing
> > we can do about the hardware: it is there, we can't change it, while we
> > can restrict ourselves to using software that is working on said
> > hardware.
> >
> > Thus, we inverse the logic, to move the condition onto the software
> > side: whenever NaN-2008 are selected, we restrict the toolchain
> > selection to at least a gcc-4.9.
> >
> > But now, the option with the NaN type is always set, so we must enclose
> > the code in gcc.mk inside a HAS_NAN_OPTION condition, as is already done
> > for the external toolchain case.
> >
> > Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> > Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> > Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
> > ---
> > arch/Config.in.mips | 2 +-
> > package/gcc/gcc.mk | 2 ++
> > 2 files changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/arch/Config.in.mips b/arch/Config.in.mips
> > index c08b2a2968..13d1a477e6 100644
> > --- a/arch/Config.in.mips
> > +++ b/arch/Config.in.mips
> > @@ -176,10 +176,10 @@ config BR2_MIPS_NAN_LEGACY
> >
> > config BR2_MIPS_NAN_2008
> > bool
> > + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
>
> This series require the previous series "arch: some require a minimal gcc
> version" adding this option.
Yes, that's why the four series were replies one to the previous, so that
they are correctly threaded in your MUA. As long as you use a decent MUA,
that is. ;-]
Regards,
Yann E. MORIN.
> Best regards,
> Romain
>
> >
> > choice
> > prompt "Target NaN"
> > - depends on BR2_TOOLCHAIN_HAS_MNAN_OPTION
> > depends on BR2_mips_32r5 || BR2_mips_64r5
> > default BR2_MIPS_ENABLE_NAN_2008
> > help
> > diff --git a/package/gcc/gcc.mk b/package/gcc/gcc.mk
> > index 49ccccf56f..d45b6d7a06 100644
> > --- a/package/gcc/gcc.mk
> > +++ b/package/gcc/gcc.mk
> > @@ -207,9 +207,11 @@ endif
> > ifneq ($(call qstrip,$(BR2_GCC_TARGET_ABI)),)
> > HOST_GCC_COMMON_CONF_OPTS += --with-abi=$(BR2_GCC_TARGET_ABI)
> > endif
> > +ifeq ($(BR2_TOOLCHAIN_HAS_MNAN_OPTION),y)
> > ifneq ($(call qstrip,$(BR2_GCC_TARGET_NAN)),)
> > HOST_GCC_COMMON_CONF_OPTS += --with-nan=$(BR2_GCC_TARGET_NAN)
> > endif
> > +endif
> > ifneq ($(call qstrip,$(BR2_GCC_TARGET_FP32_MODE)),)
> > HOST_GCC_COMMON_CONF_OPTS += --with-fp-32=$(BR2_GCC_TARGET_FP32_MODE)
> > endif
> >
>
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic
2017-10-07 12:22 ` Yann E. MORIN
@ 2017-10-07 18:43 ` Romain Naour
0 siblings, 0 replies; 51+ messages in thread
From: Romain Naour @ 2017-10-07 18:43 UTC (permalink / raw)
To: buildroot
Yann, All,
Le 07/10/2017 ? 14:22, Yann E. MORIN a ?crit?:
> Romain, All,
>
> On 2017-10-07 12:18 +0200, Romain Naour spake thusly:
>> Le 03/09/2017 ? 11:53, Yann E. MORIN a ?crit?:
>>> Currentlym the possibility to choose the NaN encoding is conditional to
>>> having a sufficiently recent gcc version.
>>>
>>> Which means that the architecture selection depends on the gcc version.
>>>
>>> But that's opposite to what we've always done in Buildroot: the software
>>> versions are conditional to the architecture options. There is nothing
>>> we can do about the hardware: it is there, we can't change it, while we
>>> can restrict ourselves to using software that is working on said
>>> hardware.
>>>
>>> Thus, we inverse the logic, to move the condition onto the software
>>> side: whenever NaN-2008 are selected, we restrict the toolchain
>>> selection to at least a gcc-4.9.
>>>
>>> But now, the option with the NaN type is always set, so we must enclose
>>> the code in gcc.mk inside a HAS_NAN_OPTION condition, as is already done
>>> for the external toolchain case.
>>>
>>> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
>>> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
>>> Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
>>> ---
>>> arch/Config.in.mips | 2 +-
>>> package/gcc/gcc.mk | 2 ++
>>> 2 files changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/arch/Config.in.mips b/arch/Config.in.mips
>>> index c08b2a2968..13d1a477e6 100644
>>> --- a/arch/Config.in.mips
>>> +++ b/arch/Config.in.mips
>>> @@ -176,10 +176,10 @@ config BR2_MIPS_NAN_LEGACY
>>>
>>> config BR2_MIPS_NAN_2008
>>> bool
>>> + select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
>>
>> This series require the previous series "arch: some require a minimal gcc
>> version" adding this option.
>
> Yes, that's why the four series were replies one to the previous, so that
> they are correctly threaded in your MUA. As long as you use a decent MUA,
> that is. ;-]
I'm only looking at the patchwork web interface and git pwc :p
Best regards,
Romain
>
> Regards,
> Yann E. MORIN.
>
>> Best regards,
>> Romain
>>
>>>
>>> choice
>>> prompt "Target NaN"
>>> - depends on BR2_TOOLCHAIN_HAS_MNAN_OPTION
>>> depends on BR2_mips_32r5 || BR2_mips_64r5
>>> default BR2_MIPS_ENABLE_NAN_2008
>>> help
>>> diff --git a/package/gcc/gcc.mk b/package/gcc/gcc.mk
>>> index 49ccccf56f..d45b6d7a06 100644
>>> --- a/package/gcc/gcc.mk
>>> +++ b/package/gcc/gcc.mk
>>> @@ -207,9 +207,11 @@ endif
>>> ifneq ($(call qstrip,$(BR2_GCC_TARGET_ABI)),)
>>> HOST_GCC_COMMON_CONF_OPTS += --with-abi=$(BR2_GCC_TARGET_ABI)
>>> endif
>>> +ifeq ($(BR2_TOOLCHAIN_HAS_MNAN_OPTION),y)
>>> ifneq ($(call qstrip,$(BR2_GCC_TARGET_NAN)),)
>>> HOST_GCC_COMMON_CONF_OPTS += --with-nan=$(BR2_GCC_TARGET_NAN)
>>> endif
>>> +endif
>>> ifneq ($(call qstrip,$(BR2_GCC_TARGET_FP32_MODE)),)
>>> HOST_GCC_COMMON_CONF_OPTS += --with-fp-32=$(BR2_GCC_TARGET_FP32_MODE)
>>> endif
>>>
>>
>
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 2/3] arch/mips: inverse the mfpxx logic
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic Yann E. MORIN
@ 2017-09-03 9:53 ` Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 3/3] toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008 Yann E. MORIN
` (2 subsequent siblings)
4 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:53 UTC (permalink / raw)
To: buildroot
Currently, the possibility to choose the floating point mode (32, xx or
64) is conditional on having a sufficiently recent gcc version.
Which means that the architecture selection depends on the gcc version.
But that's opposite to what we've always done in Buildroot: the software
versions are conditional to the architecture options. There is nothing
we can do about the hardware: it is there, we can't change it, while we
can restrict ourselves to using software that is working on said
hardware.
Thus, we inverse the logic, to move the condition onto the software
side: whenever mfpxx is selected, we restrict the toolchain selection to
at least a gcc-5.
And now, the blind BR2_TOOLCHAIN_HAS_MFPXX_OPTION symbol is no longer
needed, so we get rid of it.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
---
arch/Config.in.mips | 4 ++--
toolchain/toolchain-common.in | 4 ----
2 files changed, 2 insertions(+), 6 deletions(-)
diff --git a/arch/Config.in.mips b/arch/Config.in.mips
index 13d1a477e6..f8e57bab33 100644
--- a/arch/Config.in.mips
+++ b/arch/Config.in.mips
@@ -146,7 +146,7 @@ config BR2_MIPS_SOFT_FLOAT
choice
prompt "FP mode"
depends on !BR2_ARCH_IS_64 && !BR2_MIPS_SOFT_FLOAT
- default BR2_MIPS_FP32_MODE_XX if BR2_TOOLCHAIN_HAS_MFPXX_OPTION
+ default BR2_MIPS_FP32_MODE_XX
help
MIPS32 supports different FP modes (32,xx,64). Information about FP
modes can be found here:
@@ -159,7 +159,7 @@ config BR2_MIPS_FP32_MODE_32
config BR2_MIPS_FP32_MODE_XX
bool "xx"
- depends on BR2_TOOLCHAIN_HAS_MFPXX_OPTION
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_MIPS_FP32_MODE_64
bool "64"
diff --git a/toolchain/toolchain-common.in b/toolchain/toolchain-common.in
index dd192b9ba4..0002682e12 100644
--- a/toolchain/toolchain-common.in
+++ b/toolchain/toolchain-common.in
@@ -346,10 +346,6 @@ config BR2_TOOLCHAIN_HAS_MNAN_OPTION
bool
default y if BR2_TOOLCHAIN_GCC_AT_LEAST_4_9
-config BR2_TOOLCHAIN_HAS_MFPXX_OPTION
- bool
- default y if BR2_TOOLCHAIN_GCC_AT_LEAST_5
-
config BR2_TOOLCHAIN_HAS_SYNC_1
bool
default y
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 3/3] toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 1/3] arch/mips: inverse the NaN logic Yann E. MORIN
2017-09-03 9:53 ` [Buildroot] [PATCH 2/3] arch/mips: inverse the mfpxx logic Yann E. MORIN
@ 2017-09-03 9:53 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
2017-11-24 21:58 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Thomas Petazzoni
4 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 9:53 UTC (permalink / raw)
To: buildroot
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Romain Naour <romain.naour@openwide.fr>
Cc: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
---
toolchain/toolchain-buildroot/Config.in | 1 +
1 file changed, 1 insertion(+)
diff --git a/toolchain/toolchain-buildroot/Config.in b/toolchain/toolchain-buildroot/Config.in
index f47001fcf6..94bf1d7a09 100644
--- a/toolchain/toolchain-buildroot/Config.in
+++ b/toolchain/toolchain-buildroot/Config.in
@@ -51,6 +51,7 @@ config BR2_TOOLCHAIN_BUILDROOT_GLIBC
depends on BR2_USE_MMU
depends on !BR2_STATIC_LIBS
depends on BR2_TOOLCHAIN_HEADERS_AT_LEAST_3_2
+ depends on BR2_TOOLCHAIN_HEADERS_AT_LEAST_4_5 || !BR2_MIPS_NAN_2008
depends on !BR2_powerpc_SPE
select BR2_TOOLCHAIN_USES_GLIBC
# our glibc.mk enables RPC support
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread
* [Buildroot] [pull request] arch/arm: add some new armv8a cores
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
` (2 preceding siblings ...)
2017-09-03 9:53 ` [Buildroot] [PATCH 3/3] toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008 Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 1/9] arch/arm: re-order cores choice Yann E. MORIN
` (10 more replies)
2017-11-24 21:58 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Thomas Petazzoni
4 siblings, 11 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Hello All!
This series adds all the armv8a cores currently known to any released
version of gcc.
First, we do a bit of cleanups in the existing menu: re-order the cores,
simplify hiding non-64-bit cores, rename armv8 to armv8a.
Second, as all armv8a cores are guaranteed by the ISA to have a VFPv4,
we disallow building a soft-float toolchain on those cores when in
64-bit mode.
Now we add a bunch of armv8a cores, cortex and non-cortex alike, with a
few armv8.1a variants as well.
It is to be noted that we do have at least one armv8a core that is in
fact 32-bit only, while we also do have a few armv8a cores that are
64-bit only (or at least gcc does not recognises them in arm mode).
Eventually, we add a build-only test for all those toolchains, to check
that at least the toolchains do build. This final commit is not meant to
be applied, it adds way too many tests.
This series relies on the previous architecture series I've recently
sent:
http://lists.busybox.net/pipermail/buildroot/2017-September/201403.html
http://lists.busybox.net/pipermail/buildroot/2017-September/201409.html
Regards,
Yann E. MORIN.
PS. I've sent this series as a real pull-request, just to see how
the new patchwork handles it: currently, series are sorted by the
first patch, and patchwork misses the cover-letter. Let's see if
a cover-letter that is a pull-request gets noticed.
The following changes since commit 41eeac74f986790fe26f03ebf05322120d2c04da
package/gcc: slight cleanup and reorg in remaining arch depends (2017-09-03 11:14:53 +0200)
are available in the git repository at:
git://git.buildroot.org/~ymorin/git/buildroot.git
for you to fetch changes up to a876f6b64754ba3b96492c57b18654f8ec465ca0
[DON'T COMMIT] tests for all new arm cores (2017-09-03 15:02:47 +0200)
----------------------------------------------------------------
Yann E. MORIN (9):
arch/arm: re-order cores choice
arch/arm: simplify hiding non 64-bit cores
arch/arm: armv8 is really armv8a
arch/arm: do not allow soft-float for armv8a
arch/arm: add cortex-A32
arch/arm: add some armv8a cortex variants
arch/arm: add some non-cortex armv8a cores
arch/arm: add armv8.1a cores
[DON'T COMMIT] tests for all new arm cores
arch/Config.in.arm | 297 ++++++++++++++++-----
package/pkg-cmake.mk | 2 +-
.../tests/toolchain/test_internal_aarch64_all.py | 162 +++++++++++
.../tests/toolchain/test_internal_arm_all.py | 95 +++++++
.../toolchain-external-codesourcery-arm/Config.in | 2 +-
.../toolchain-external-linaro-arm/Config.in | 2 +-
.../toolchain-external-linaro-armeb/Config.in | 2 +-
7 files changed, 489 insertions(+), 73 deletions(-)
create mode 100644 support/testing/tests/toolchain/test_internal_aarch64_all.py
create mode 100644 support/testing/tests/toolchain/test_internal_arm_all.py
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 1/9] arch/arm: re-order cores choice
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 2/9] arch/arm: simplify hiding non 64-bit cores Yann E. MORIN
` (9 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Currently, the logic for ordering the ARM cores in the choice is all but
obvious. ;-)
Reorder the choice by architecture generation, starting with armv4,
ending with armv8.
Add a comment before each generation, just for ease of use. Add a
separate comment for armv7a and armv7m.
Finally, order cores alphabetically inside the same generation (except
for armv7m cores, listed after all armv7a cores.)
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 116 +++++++++++++++++++++++++++++++----------------------
1 file changed, 69 insertions(+), 47 deletions(-)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index b58be6bce0..66dce3a562 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -68,6 +68,8 @@ choice
help
Specific CPU variant to use
+comment "armv4 cores"
+ depends on !BR2_ARCH_IS_64
config BR2_arm920t
bool "arm920t"
select BR2_ARM_CPU_HAS_ARM
@@ -82,6 +84,21 @@ config BR2_arm922t
select BR2_ARM_CPU_ARMV4
select BR2_ARCH_HAS_MMU_OPTIONAL
depends on !BR2_ARCH_IS_64
+config BR2_fa526
+ bool "fa526/626"
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_ARMV4
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ depends on !BR2_ARCH_IS_64
+config BR2_strongarm
+ bool "strongarm sa110/sa1100"
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_ARMV4
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ depends on !BR2_ARCH_IS_64
+
+comment "armv5 cores"
+ depends on !BR2_ARCH_IS_64
config BR2_arm926t
bool "arm926t"
select BR2_ARM_CPU_HAS_ARM
@@ -90,6 +107,22 @@ config BR2_arm926t
select BR2_ARM_CPU_ARMV5
select BR2_ARCH_HAS_MMU_OPTIONAL
depends on !BR2_ARCH_IS_64
+config BR2_iwmmxt
+ bool "iwmmxt"
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_ARMV5
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ depends on !BR2_ARCH_IS_64
+config BR2_xscale
+ bool "xscale"
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_HAS_THUMB
+ select BR2_ARM_CPU_ARMV5
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ depends on !BR2_ARCH_IS_64
+
+comment "armv6 cores"
+ depends on !BR2_ARCH_IS_64
config BR2_arm1136j_s
bool "arm1136j-s"
select BR2_ARM_CPU_HAS_ARM
@@ -128,6 +161,9 @@ config BR2_arm11mpcore
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
depends on !BR2_ARCH_IS_64
+
+comment "armv7a cores"
+ depends on !BR2_ARCH_IS_64
config BR2_cortex_a5
bool "cortex-A5"
select BR2_ARM_CPU_HAS_ARM
@@ -212,6 +248,28 @@ config BR2_cortex_a17_a7
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
depends on !BR2_ARCH_IS_64
+config BR2_pj4
+ bool "pj4"
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_HAS_VFPV3
+ select BR2_ARM_CPU_ARMV7A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ depends on !BR2_ARCH_IS_64
+
+comment "armv7m cores"
+ depends on !BR2_ARCH_IS_64
+config BR2_cortex_m3
+ bool "cortex-M3"
+ select BR2_ARM_CPU_HAS_THUMB2
+ select BR2_ARM_CPU_ARMV7M
+ depends on !BR2_ARCH_IS_64
+config BR2_cortex_m4
+ bool "cortex-M4"
+ select BR2_ARM_CPU_HAS_THUMB2
+ select BR2_ARM_CPU_ARMV7M
+ depends on !BR2_ARCH_IS_64
+
+comment "armv8 cores"
config BR2_cortex_a53
bool "cortex-A53"
select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
@@ -255,48 +313,6 @@ config BR2_cortex_a72_a53
select BR2_ARM_CPU_ARMV8
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
-config BR2_cortex_m3
- bool "cortex-M3"
- select BR2_ARM_CPU_HAS_THUMB2
- select BR2_ARM_CPU_ARMV7M
- depends on !BR2_ARCH_IS_64
-config BR2_cortex_m4
- bool "cortex-M4"
- select BR2_ARM_CPU_HAS_THUMB2
- select BR2_ARM_CPU_ARMV7M
- depends on !BR2_ARCH_IS_64
-config BR2_fa526
- bool "fa526/626"
- select BR2_ARM_CPU_HAS_ARM
- select BR2_ARM_CPU_ARMV4
- select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
-config BR2_pj4
- bool "pj4"
- select BR2_ARM_CPU_HAS_ARM
- select BR2_ARM_CPU_HAS_VFPV3
- select BR2_ARM_CPU_ARMV7A
- select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
-config BR2_strongarm
- bool "strongarm sa110/sa1100"
- select BR2_ARM_CPU_HAS_ARM
- select BR2_ARM_CPU_ARMV4
- select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
-config BR2_xscale
- bool "xscale"
- select BR2_ARM_CPU_HAS_ARM
- select BR2_ARM_CPU_HAS_THUMB
- select BR2_ARM_CPU_ARMV5
- select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
-config BR2_iwmmxt
- bool "iwmmxt"
- select BR2_ARM_CPU_HAS_ARM
- select BR2_ARM_CPU_ARMV5
- select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
endchoice
config BR2_ARM_ENABLE_NEON
@@ -551,15 +567,23 @@ config BR2_ENDIAN
default "BIG" if (BR2_armeb || BR2_aarch64_be)
config BR2_GCC_TARGET_CPU
+ # armv4
default "arm920t" if BR2_arm920t
default "arm922t" if BR2_arm922t
+ default "fa526" if BR2_fa526
+ default "strongarm" if BR2_strongarm
+ # armv5
default "arm926ej-s" if BR2_arm926t
+ default "iwmmxt" if BR2_iwmmxt
+ default "xscale" if BR2_xscale
+ # armv6
default "arm1136j-s" if BR2_arm1136j_s
default "arm1136jf-s" if BR2_arm1136jf_s
default "arm1176jz-s" if BR2_arm1176jz_s
default "arm1176jzf-s" if BR2_arm1176jzf_s
default "mpcore" if BR2_arm11mpcore && BR2_ARM_CPU_HAS_VFPV2
default "mpcorenovfp" if BR2_arm11mpcore
+ # armv7a
default "cortex-a5" if BR2_cortex_a5
default "cortex-a7" if BR2_cortex_a7
default "cortex-a8" if BR2_cortex_a8
@@ -569,13 +593,11 @@ config BR2_GCC_TARGET_CPU
default "cortex-a15.cortex-a7" if BR2_cortex_a15_a7
default "cortex-a17" if BR2_cortex_a17
default "cortex-a17.cortex-a7" if BR2_cortex_a17_a7
+ default "marvell-pj4" if BR2_pj4
+ # armv7m
default "cortex-m3" if BR2_cortex_m3
default "cortex-m4" if BR2_cortex_m4
- default "fa526" if BR2_fa526
- default "marvell-pj4" if BR2_pj4
- default "strongarm" if BR2_strongarm
- default "xscale" if BR2_xscale
- default "iwmmxt" if BR2_iwmmxt
+ # armv8
default "cortex-a53" if BR2_cortex_a53
default "cortex-a57" if BR2_cortex_a57
default "cortex-a57.cortex-a53" if BR2_cortex_a57_a53
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 2/9] arch/arm: simplify hiding non 64-bit cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 1/9] arch/arm: re-order cores choice Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 3/9] arch/arm: armv8 is really armv8a Yann E. MORIN
` (8 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Now that the cores are all oredered correctly, we can just enclose all
the non 64-bit cores inside a big if-block, rather than have each of
them have the dependency.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 31 ++-----------------------------
1 file changed, 2 insertions(+), 29 deletions(-)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 66dce3a562..c2b7931528 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -68,37 +68,32 @@ choice
help
Specific CPU variant to use
+if !BR2_ARCH_IS_64
comment "armv4 cores"
- depends on !BR2_ARCH_IS_64
config BR2_arm920t
bool "arm920t"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV4
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_arm922t
bool "arm922t"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV4
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_fa526
bool "fa526/626"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_ARMV4
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_strongarm
bool "strongarm sa110/sa1100"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_ARMV4
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
comment "armv5 cores"
- depends on !BR2_ARCH_IS_64
config BR2_arm926t
bool "arm926t"
select BR2_ARM_CPU_HAS_ARM
@@ -106,30 +101,25 @@ config BR2_arm926t
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV5
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_iwmmxt
bool "iwmmxt"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_ARMV5
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_xscale
bool "xscale"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV5
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
comment "armv6 cores"
- depends on !BR2_ARCH_IS_64
config BR2_arm1136j_s
bool "arm1136j-s"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_arm1136jf_s
bool "arm1136jf-s"
select BR2_ARM_CPU_HAS_ARM
@@ -137,14 +127,12 @@ config BR2_arm1136jf_s
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_arm1176jz_s
bool "arm1176jz-s"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_arm1176jzf_s
bool "arm1176jzf-s"
select BR2_ARM_CPU_HAS_ARM
@@ -152,7 +140,6 @@ config BR2_arm1176jzf_s
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_arm11mpcore
bool "mpcore"
select BR2_ARM_CPU_HAS_ARM
@@ -160,10 +147,8 @@ config BR2_arm11mpcore
select BR2_ARM_CPU_HAS_THUMB
select BR2_ARM_CPU_ARMV6
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
comment "armv7a cores"
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a5
bool "cortex-A5"
select BR2_ARM_CPU_HAS_ARM
@@ -172,7 +157,6 @@ config BR2_cortex_a5
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a7
bool "cortex-A7"
select BR2_ARM_CPU_HAS_ARM
@@ -181,7 +165,6 @@ config BR2_cortex_a7
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a8
bool "cortex-A8"
select BR2_ARM_CPU_HAS_ARM
@@ -190,7 +173,6 @@ config BR2_cortex_a8
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a9
bool "cortex-A9"
select BR2_ARM_CPU_HAS_ARM
@@ -199,7 +181,6 @@ config BR2_cortex_a9
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a12
bool "cortex-A12"
select BR2_ARM_CPU_HAS_ARM
@@ -208,7 +189,6 @@ config BR2_cortex_a12
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a15
bool "cortex-A15"
select BR2_ARM_CPU_HAS_ARM
@@ -217,7 +197,6 @@ config BR2_cortex_a15
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a15_a7
bool "cortex-A15/A7 big.LITTLE"
select BR2_ARM_CPU_HAS_ARM
@@ -227,7 +206,6 @@ config BR2_cortex_a15_a7
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_4_9
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a17
bool "cortex-A17"
select BR2_ARM_CPU_HAS_ARM
@@ -237,7 +215,6 @@ config BR2_cortex_a17
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
- depends on !BR2_ARCH_IS_64
config BR2_cortex_a17_a7
bool "cortex-A17/A7 big.LITTLE"
select BR2_ARM_CPU_HAS_ARM
@@ -247,27 +224,23 @@ config BR2_cortex_a17_a7
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
- depends on !BR2_ARCH_IS_64
config BR2_pj4
bool "pj4"
select BR2_ARM_CPU_HAS_ARM
select BR2_ARM_CPU_HAS_VFPV3
select BR2_ARM_CPU_ARMV7A
select BR2_ARCH_HAS_MMU_OPTIONAL
- depends on !BR2_ARCH_IS_64
comment "armv7m cores"
- depends on !BR2_ARCH_IS_64
config BR2_cortex_m3
bool "cortex-M3"
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7M
- depends on !BR2_ARCH_IS_64
config BR2_cortex_m4
bool "cortex-M4"
select BR2_ARM_CPU_HAS_THUMB2
select BR2_ARM_CPU_ARMV7M
- depends on !BR2_ARCH_IS_64
+endif # !BR2_ARCH_IS_64
comment "armv8 cores"
config BR2_cortex_a53
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 3/9] arch/arm: armv8 is really armv8a
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 1/9] arch/arm: re-order cores choice Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 2/9] arch/arm: simplify hiding non 64-bit cores Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a Yann E. MORIN
` (7 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
For armv8, there are different profiles: A, M and R, like there is for
armv7.
So, rename our internal symbol to mirror what we do for armv7.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 14 +++++++-------
package/pkg-cmake.mk | 2 +-
.../toolchain-external-codesourcery-arm/Config.in | 2 +-
.../toolchain-external-linaro-arm/Config.in | 2 +-
.../toolchain-external-linaro-armeb/Config.in | 2 +-
5 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index c2b7931528..d110d52e2d 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -59,7 +59,7 @@ config BR2_ARM_CPU_ARMV7A
config BR2_ARM_CPU_ARMV7M
bool
-config BR2_ARM_CPU_ARMV8
+config BR2_ARM_CPU_ARMV8A
bool
choice
@@ -249,7 +249,7 @@ config BR2_cortex_a53
select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_FP_ARMV8
- select BR2_ARM_CPU_ARMV8
+ select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
config BR2_cortex_a57
bool "cortex-A57"
@@ -257,7 +257,7 @@ config BR2_cortex_a57
select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_FP_ARMV8
- select BR2_ARM_CPU_ARMV8
+ select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
config BR2_cortex_a57_a53
bool "cortex-A57/A53 big.LITTLE"
@@ -265,7 +265,7 @@ config BR2_cortex_a57_a53
select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_FP_ARMV8
- select BR2_ARM_CPU_ARMV8
+ select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_cortex_a72
@@ -274,7 +274,7 @@ config BR2_cortex_a72
select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_FP_ARMV8
- select BR2_ARM_CPU_ARMV8
+ select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
config BR2_cortex_a72_a53
@@ -283,7 +283,7 @@ config BR2_cortex_a72_a53
select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
select BR2_ARM_CPU_HAS_FP_ARMV8
- select BR2_ARM_CPU_ARMV8
+ select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
endchoice
@@ -570,7 +570,7 @@ config BR2_GCC_TARGET_CPU
# armv7m
default "cortex-m3" if BR2_cortex_m3
default "cortex-m4" if BR2_cortex_m4
- # armv8
+ # armv8a
default "cortex-a53" if BR2_cortex_a53
default "cortex-a57" if BR2_cortex_a57
default "cortex-a57.cortex-a53" if BR2_cortex_a57_a53
diff --git a/package/pkg-cmake.mk b/package/pkg-cmake.mk
index db78d897d8..6739704e3c 100644
--- a/package/pkg-cmake.mk
+++ b/package/pkg-cmake.mk
@@ -224,7 +224,7 @@ else ifeq ($(BR2_ARM_CPU_ARMV6),y)
CMAKE_SYSTEM_PROCESSOR_ARM_VARIANT = armv6
else ifeq ($(BR2_ARM_CPU_ARMV7A),y)
CMAKE_SYSTEM_PROCESSOR_ARM_VARIANT = armv7
-else ifeq ($(BR2_ARM_CPU_ARMV8),y)
+else ifeq ($(BR2_ARM_CPU_ARMV8A),y)
CMAKE_SYSTEM_PROCESSOR_ARM_VARIANT = armv8
endif
diff --git a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
index 60d86c33d5..bdbe2acaea 100644
--- a/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-codesourcery-arm/Config.in
@@ -5,7 +5,7 @@ config BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_ARM
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABI
# Unsupported ARM cores
- depends on !BR2_cortex_a12 && !BR2_cortex_a17 && !BR2_ARM_CPU_ARMV8
+ depends on !BR2_cortex_a12 && !BR2_cortex_a17 && !BR2_ARM_CPU_ARMV8A
depends on !BR2_STATIC_LIBS
select BR2_TOOLCHAIN_EXTERNAL_GLIBC
select BR2_TOOLCHAIN_HAS_SSP
diff --git a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
index f5bbbaaad8..51604d9dde 100644
--- a/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-linaro-arm/Config.in
@@ -8,7 +8,7 @@ config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARM
bool "Linaro ARM 2017.02"
depends on BR2_arm
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
- depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
+ depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8A
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABIHF
depends on !BR2_STATIC_LIBS
diff --git a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
index 64e97eeea7..2f01f6994f 100644
--- a/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
+++ b/toolchain/toolchain-external/toolchain-external-linaro-armeb/Config.in
@@ -8,7 +8,7 @@ config BR2_TOOLCHAIN_EXTERNAL_LINARO_ARMEB
bool "Linaro armeb 2017.02"
depends on BR2_armeb
depends on !BR2_ARCH_NEEDS_GCC_AT_LEAST_7
- depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8
+ depends on BR2_ARM_CPU_ARMV7A || BR2_ARM_CPU_ARMV8A
depends on BR2_HOSTARCH = "x86_64" || BR2_HOSTARCH = "x86"
depends on BR2_ARM_EABIHF
depends on !BR2_STATIC_LIBS
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (2 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 3/9] arch/arm: armv8 is really armv8a Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-04 17:21 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 5/9] arch/arm: add cortex-A32 Yann E. MORIN
` (6 subsequent siblings)
10 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
armv8a has made VFPv4 and NEON mandatory, so there is no point in
allowing software floating point, even in 32-bit mode.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 1 +
1 file changed, 1 insertion(+)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index d110d52e2d..0ae6a16a0a 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -374,6 +374,7 @@ choice
config BR2_ARM_SOFT_FLOAT
bool "Soft float"
depends on BR2_ARM_EABI
+ depends on !BR2_ARM_CPU_ARMV8A
select BR2_SOFT_FLOAT
help
This option allows to use software emulated floating
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-09-03 13:17 ` [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a Yann E. MORIN
@ 2017-09-04 17:21 ` Yann E. MORIN
2017-11-24 22:08 ` Thomas Petazzoni
0 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-04 17:21 UTC (permalink / raw)
To: buildroot
Thomas, All,
On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
> armv8a has made VFPv4 and NEON mandatory, so there is no point in
> allowing software floating point, even in 32-bit mode.
In fact, even though I'm pretty sure that OK in 64-bit, we might still
have a reason to do soft-float in 32-bit mode, if only for legacy
binary-only applications that were built as soft-float way back in the
(not so) good old days...
In which case, we should instead depend on !BR2_ARCH_IS_64.
Thoughts?
Regards,
Yann E. MORIN.
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> ---
> arch/Config.in.arm | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/arch/Config.in.arm b/arch/Config.in.arm
> index d110d52e2d..0ae6a16a0a 100644
> --- a/arch/Config.in.arm
> +++ b/arch/Config.in.arm
> @@ -374,6 +374,7 @@ choice
> config BR2_ARM_SOFT_FLOAT
> bool "Soft float"
> depends on BR2_ARM_EABI
> + depends on !BR2_ARM_CPU_ARMV8A
> select BR2_SOFT_FLOAT
> help
> This option allows to use software emulated floating
> --
> 2.11.0
>
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-09-04 17:21 ` Yann E. MORIN
@ 2017-11-24 22:08 ` Thomas Petazzoni
2017-11-25 17:10 ` Arnout Vandecappelle
0 siblings, 1 reply; 51+ messages in thread
From: Thomas Petazzoni @ 2017-11-24 22:08 UTC (permalink / raw)
To: buildroot
Hello,
On Mon, 4 Sep 2017 19:21:37 +0200, Yann E. MORIN wrote:
> Thomas, All,
>
> On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
> > armv8a has made VFPv4 and NEON mandatory, so there is no point in
> > allowing software floating point, even in 32-bit mode.
>
> In fact, even though I'm pretty sure that OK in 64-bit, we might still
> have a reason to do soft-float in 32-bit mode, if only for legacy
> binary-only applications that were built as soft-float way back in the
> (not so) good old days...
>
> In which case, we should instead depend on !BR2_ARCH_IS_64.
>
> Thoughts?
I'm hesitating on this one. I believe the use-case you mention makes
sense, however I believe that it can still be supported with your
proposed patch.
Indeed, even if you want to run a binary-only soft-float application,
it doesn't prevent the rest of your system from using VFP, as long as
you use EABI (and not EABIhf).
The whole reason why EABI passes floating-point arguments into integer
registers is precisely to allow compatibility between soft-float code
and hard-float code. Thanks to that, a soft-float function can call a
hard-float function, and vice-versa.
Therefore, even if you have a binary-only soft-float application, there
is really no reason to build the entire system soft-float.
So, I think your original patch is OK as-is. However, I still see one
inconsistent thing: if we disallow soft-float because ARMv8-A mandates
VFPv4, why would we allow soft-float for ARMv7-A cores that have a
VFP ? To be consistent, we should also disallow soft-float.
Another way to think about this is: is it possible to build an AArch64
soft-float system? do we want to allow the user to select what is
*possible* or what makes sense?
Best regards,
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-11-24 22:08 ` Thomas Petazzoni
@ 2017-11-25 17:10 ` Arnout Vandecappelle
2017-11-26 11:12 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Arnout Vandecappelle @ 2017-11-25 17:10 UTC (permalink / raw)
To: buildroot
On 24-11-17 23:08, Thomas Petazzoni wrote:
> Hello,
>
> On Mon, 4 Sep 2017 19:21:37 +0200, Yann E. MORIN wrote:
>> Thomas, All,
>>
>> On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
>>> armv8a has made VFPv4 and NEON mandatory, so there is no point in
>>> allowing software floating point, even in 32-bit mode.
>>
>> In fact, even though I'm pretty sure that OK in 64-bit, we might still
>> have a reason to do soft-float in 32-bit mode, if only for legacy
>> binary-only applications that were built as soft-float way back in the
>> (not so) good old days...
>>
>> In which case, we should instead depend on !BR2_ARCH_IS_64.
>>
>> Thoughts?
>
> I'm hesitating on this one. I believe the use-case you mention makes
> sense, however I believe that it can still be supported with your
> proposed patch.
>
> Indeed, even if you want to run a binary-only soft-float application,
> it doesn't prevent the rest of your system from using VFP, as long as
> you use EABI (and not EABIhf).
>
> The whole reason why EABI passes floating-point arguments into integer
> registers is precisely to allow compatibility between soft-float code
> and hard-float code. Thanks to that, a soft-float function can call a
> hard-float function, and vice-versa.
>
> Therefore, even if you have a binary-only soft-float application, there
> is really no reason to build the entire system soft-float.
>
> So, I think your original patch is OK as-is. However, I still see one
> inconsistent thing: if we disallow soft-float because ARMv8-A mandates
> VFPv4, why would we allow soft-float for ARMv7-A cores that have a
> VFP ? To be consistent, we should also disallow soft-float.
>
> Another way to think about this is: is it possible to build an AArch64
> soft-float system? do we want to allow the user to select what is
> *possible* or what makes sense?
gcc aarch64 doesn't have a float-abi option, so no it's not possible to build
with soft-float. So the depends on BR2_ARM_EABI is certainly needed.
That said, I don't think it makes a lot of sense to add extra 'depends on'
lines just to avoid that the user would select an option that probably doesn't
make sense. We already make sure that the default is appropriate. Let's not make
our life more difficult.
So I'd say: reject this patch.
Regards,
Arnout
--
Arnout Vandecappelle arnout at mind be
Senior Embedded Software Architect +32-16-286500
Essensium/Mind http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint: 7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-11-25 17:10 ` Arnout Vandecappelle
@ 2017-11-26 11:12 ` Yann E. MORIN
2017-11-26 11:18 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-11-26 11:12 UTC (permalink / raw)
To: buildroot
Arnout, All,
On 2017-11-25 18:10 +0100, Arnout Vandecappelle spake thusly:
> On 24-11-17 23:08, Thomas Petazzoni wrote:
> > Hello,
> >
> > On Mon, 4 Sep 2017 19:21:37 +0200, Yann E. MORIN wrote:
> >> Thomas, All,
> >>
> >> On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
> >>> armv8a has made VFPv4 and NEON mandatory, so there is no point in
> >>> allowing software floating point, even in 32-bit mode.
> >>
> >> In fact, even though I'm pretty sure that OK in 64-bit, we might still
> >> have a reason to do soft-float in 32-bit mode, if only for legacy
> >> binary-only applications that were built as soft-float way back in the
> >> (not so) good old days...
> >>
> >> In which case, we should instead depend on !BR2_ARCH_IS_64.
> >>
> >> Thoughts?
> >
> > I'm hesitating on this one. I believe the use-case you mention makes
> > sense, however I believe that it can still be supported with your
> > proposed patch.
> >
> > Indeed, even if you want to run a binary-only soft-float application,
> > it doesn't prevent the rest of your system from using VFP, as long as
> > you use EABI (and not EABIhf).
> >
> > The whole reason why EABI passes floating-point arguments into integer
> > registers is precisely to allow compatibility between soft-float code
> > and hard-float code. Thanks to that, a soft-float function can call a
> > hard-float function, and vice-versa.
> >
> > Therefore, even if you have a binary-only soft-float application, there
> > is really no reason to build the entire system soft-float.
> >
> > So, I think your original patch is OK as-is. However, I still see one
> > inconsistent thing: if we disallow soft-float because ARMv8-A mandates
> > VFPv4, why would we allow soft-float for ARMv7-A cores that have a
> > VFP ? To be consistent, we should also disallow soft-float.
> >
> > Another way to think about this is: is it possible to build an AArch64
> > soft-float system? do we want to allow the user to select what is
> > *possible* or what makes sense?
>
> gcc aarch64
It's not about aarch64, it's about aarch32, i.e. an armv8a core running
in 32-bit mode.
> doesn't have a float-abi option, so no it's not possible to build
> with soft-float. So the depends on BR2_ARM_EABI is certainly needed.
>
> That said, I don't think it makes a lot of sense to add extra 'depends on'
> lines just to avoid that the user would select an option that probably doesn't
> make sense. We already make sure that the default is appropriate. Let's not make
> our life more difficult.
It's not about making *our* lives more dificult or easier. It's about
making the user's lives easier.
If there is a prompt that offers an option, but that has no impact on
the resulting output, then that prompt should not exist. Otherwise, it
is very confusing.
Regards,
Yann E. MORIN.
> So I'd say: reject this patch.
>
> Regards,
> Arnout
>
>
> --
> Arnout Vandecappelle arnout at mind be
> Senior Embedded Software Architect +32-16-286500
> Essensium/Mind http://www.mind.be
> G.Geenslaan 9, 3001 Leuven, Belgium BE 872 984 063 RPR Leuven
> LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
> GPG fingerprint: 7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a
2017-11-26 11:12 ` Yann E. MORIN
@ 2017-11-26 11:18 ` Yann E. MORIN
0 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-11-26 11:18 UTC (permalink / raw)
To: buildroot
Arnout, Thomas, All,
On 2017-11-26 12:12 +0100, Yann E. MORIN spake thusly:
> On 2017-11-25 18:10 +0100, Arnout Vandecappelle spake thusly:
> > On 24-11-17 23:08, Thomas Petazzoni wrote:
> > > On Mon, 4 Sep 2017 19:21:37 +0200, Yann E. MORIN wrote:
> > >> Thomas, All,
> > >>
> > >> On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
> > >>> armv8a has made VFPv4 and NEON mandatory, so there is no point in
> > >>> allowing software floating point, even in 32-bit mode.
> > >>
> > >> In fact, even though I'm pretty sure that OK in 64-bit, we might still
> > >> have a reason to do soft-float in 32-bit mode, if only for legacy
> > >> binary-only applications that were built as soft-float way back in the
> > >> (not so) good old days...
> > >>
> > >> In which case, we should instead depend on !BR2_ARCH_IS_64.
> > >>
> > >> Thoughts?
> > >
> > > I'm hesitating on this one. I believe the use-case you mention makes
> > > sense, however I believe that it can still be supported with your
> > > proposed patch.
> > >
> > > Indeed, even if you want to run a binary-only soft-float application,
> > > it doesn't prevent the rest of your system from using VFP, as long as
> > > you use EABI (and not EABIhf).
> > >
> > > The whole reason why EABI passes floating-point arguments into integer
> > > registers is precisely to allow compatibility between soft-float code
> > > and hard-float code. Thanks to that, a soft-float function can call a
> > > hard-float function, and vice-versa.
> > >
> > > Therefore, even if you have a binary-only soft-float application, there
> > > is really no reason to build the entire system soft-float.
> > >
> > > So, I think your original patch is OK as-is. However, I still see one
> > > inconsistent thing: if we disallow soft-float because ARMv8-A mandates
> > > VFPv4, why would we allow soft-float for ARMv7-A cores that have a
> > > VFP ? To be consistent, we should also disallow soft-float.
> > >
> > > Another way to think about this is: is it possible to build an AArch64
> > > soft-float system? do we want to allow the user to select what is
> > > *possible* or what makes sense?
> >
> > gcc aarch64
>
> It's not about aarch64, it's about aarch32, i.e. an armv8a core running
> in 32-bit mode.
>
> > doesn't have a float-abi option, so no it's not possible to build
> > with soft-float. So the depends on BR2_ARM_EABI is certainly needed.
> >
> > That said, I don't think it makes a lot of sense to add extra 'depends on'
> > lines just to avoid that the user would select an option that probably doesn't
> > make sense. We already make sure that the default is appropriate. Let's not make
> > our life more difficult.
>
> It's not about making *our* lives more dificult or easier. It's about
> making the user's lives easier.
>
> If there is a prompt that offers an option, but that has no impact on
> the resulting output, then that prompt should not exist. Otherwise, it
> is very confusing.
Damned, I hadn't finished my mail...
> > So I'd say: reject this patch.
But seeing the argument from Thomas, yes it does not make sense to hide
soft-float anyway.
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 5/9] arch/arm: add cortex-A32
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (3 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 4/9] arch/arm: do not allow soft-float for armv8a Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 6/9] arch/arm: add some armv8a cortex variants Yann E. MORIN
` (5 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
The cortex-A32 is an armv8a core, but it lacks the optional AArch64
extensions, so can only work in 32-bit mode.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 0ae6a16a0a..0304929ba1 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -243,6 +243,16 @@ config BR2_cortex_m4
endif # !BR2_ARCH_IS_64
comment "armv8 cores"
+config BR2_cortex_a32
+ bool "cortex-A32"
+ depends on !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_ARM
+ select BR2_ARM_CPU_HAS_NEON
+ select BR2_ARM_CPU_HAS_THUMB2
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_cortex_a53
bool "cortex-A53"
select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
@@ -572,6 +582,7 @@ config BR2_GCC_TARGET_CPU
default "cortex-m3" if BR2_cortex_m3
default "cortex-m4" if BR2_cortex_m4
# armv8a
+ default "cortex-a32" if BR2_cortex_a32
default "cortex-a53" if BR2_cortex_a53
default "cortex-a57" if BR2_cortex_a57
default "cortex-a57.cortex-a53" if BR2_cortex_a57_a53
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 6/9] arch/arm: add some armv8a cortex variants
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (4 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 5/9] arch/arm: add cortex-A32 Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores Yann E. MORIN
` (4 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 40 ++++++++++++++++++++++++++++++++++++++++
1 file changed, 40 insertions(+)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 0304929ba1..65caf7aa80 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -253,6 +253,15 @@ config BR2_cortex_a32
select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+config BR2_cortex_a35
+ bool "cortex-A35"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
config BR2_cortex_a53
bool "cortex-A53"
select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
@@ -296,6 +305,33 @@ config BR2_cortex_a72_a53
select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+config BR2_cortex_a73
+ bool "cortex-A73"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_cortex_a73_a35
+ bool "cortex-A73/A35 big.LITTLE"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_cortex_a73_a53
+ bool "cortex-A73/A53 big.LITTLE"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
endchoice
config BR2_ARM_ENABLE_NEON
@@ -583,11 +619,15 @@ config BR2_GCC_TARGET_CPU
default "cortex-m4" if BR2_cortex_m4
# armv8a
default "cortex-a32" if BR2_cortex_a32
+ default "cortex-a35" if BR2_cortex_a35
default "cortex-a53" if BR2_cortex_a53
default "cortex-a57" if BR2_cortex_a57
default "cortex-a57.cortex-a53" if BR2_cortex_a57_a53
default "cortex-a72" if BR2_cortex_a72
default "cortex-a72.cortex-a53" if BR2_cortex_a72_a53
+ default "cortex-a73" if BR2_cortex_a73
+ default "cortex-a73.cortex-a35" if BR2_cortex_a73_a35
+ default "cortex-a73.cortex-a53" if BR2_cortex_a73_a53
config BR2_GCC_TARGET_ABI
default "aapcs-linux" if BR2_arm || BR2_armeb
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (5 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 6/9] arch/arm: add some armv8a cortex variants Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 14:04 ` Thomas Petazzoni
2017-09-03 13:17 ` [Buildroot] [PATCH 8/9] arch/arm: add armv8.1a cores Yann E. MORIN
` (3 subsequent siblings)
10 siblings, 1 reply; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Some need gcc-5, some gcc-6 and some gcc-7.
The thunderx familly does not build in 32-bit mode (gcc complains
that the CPU is unknown, and even gcc master only knows them as
aarch64-only).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 77 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 77 insertions(+)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 65caf7aa80..40d6ae18b8 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -332,6 +332,74 @@ config BR2_cortex_a73_a53
select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_exynos_m1
+ bool "exynos-m1"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+config BR2_falkor
+ bool "falkor"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_qdf24xx
+ bool "qdf24xx"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_6
+if BR2_ARCH_IS_64
+config BR2_thunderx
+ bool "thunderx"
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+config BR2_thunderxt81
+ bool "thunderxt81"
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_thunderxt83
+ bool "thunderxt83"
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_thunderxt88
+ bool "thunderxt88"
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_thunderxt88p1
+ bool "thunderxt88p1"
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+endif # BR2_ARCH_IS_64
+config BR2_xgene1
+ bool "xgene1"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
endchoice
config BR2_ARM_ENABLE_NEON
@@ -628,6 +696,15 @@ config BR2_GCC_TARGET_CPU
default "cortex-a73" if BR2_cortex_a73
default "cortex-a73.cortex-a35" if BR2_cortex_a73_a35
default "cortex-a73.cortex-a53" if BR2_cortex_a73_a53
+ default "exynos-m1" if BR2_exynos_m1
+ default "falkor" if BR2_falkor
+ default "qdf24xx" if BR2_qdf24xx
+ default "thunderx" if BR2_thunderx
+ default "thunderxt81" if BR2_thunderxt81
+ default "thunderxt83" if BR2_thunderxt83
+ default "thunderxt88" if BR2_thunderxt88
+ default "thunderxt88p1" if BR2_thunderxt88p1
+ default "xgene1" if BR2_xgene1
config BR2_GCC_TARGET_ABI
default "aapcs-linux" if BR2_arm || BR2_armeb
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores
2017-09-03 13:17 ` [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores Yann E. MORIN
@ 2017-09-03 14:04 ` Thomas Petazzoni
2017-09-03 15:16 ` Yann E. MORIN
0 siblings, 1 reply; 51+ messages in thread
From: Thomas Petazzoni @ 2017-09-03 14:04 UTC (permalink / raw)
To: buildroot
Hello,
On Sun, 3 Sep 2017 15:17:47 +0200, Yann E. MORIN wrote:
> Some need gcc-5, some gcc-6 and some gcc-7.
>
> The thunderx familly does not build in 32-bit mode (gcc complains
> that the CPU is unknown, and even gcc master only knows them as
> aarch64-only).
>
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> ---
> arch/Config.in.arm | 77 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 77 insertions(+)
Series looks good, but on this one, I'm wondering: do we really want to
add this long list of CPU cores, that most likely nobody is ever going
to use with Buildroot ?
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores
2017-09-03 14:04 ` Thomas Petazzoni
@ 2017-09-03 15:16 ` Yann E. MORIN
0 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 15:16 UTC (permalink / raw)
To: buildroot
Thomas, All,
On 2017-09-03 16:04 +0200, Thomas Petazzoni spake thusly:
> On Sun, 3 Sep 2017 15:17:47 +0200, Yann E. MORIN wrote:
> > Some need gcc-5, some gcc-6 and some gcc-7.
> >
> > The thunderx familly does not build in 32-bit mode (gcc complains
> > that the CPU is unknown, and even gcc master only knows them as
> > aarch64-only).
> >
> > Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> > Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> > ---
> > arch/Config.in.arm | 77 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 77 insertions(+)
>
> Series looks good, but on this one, I'm wondering: do we really want to
> add this long list of CPU cores, that most likely nobody is ever going
> to use with Buildroot ?
Well, I'm not sure, but I would not bet either...
I know for sure that some people are using Buildroot to build small
systems that they use as docker roots for example (they've spoken so
either on the list or on the IRC channel some time ago). Those are
mostly expected to run on server-class machines. The Thunder familly
of CPUs are just meant for those kind of server-class machines.
So I do see a direct use-case for adding the ThunderX familly of cores,
yes.
For the others, I'm not as positive, but I still added them while I was
at it. Not much additional work, except waiting for more tests to run...
;-)
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 8/9] arch/arm: add armv8.1a cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (6 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 7/9] arch/arm: add some non-cortex armv8a cores Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:17 ` [Buildroot] [PATCH 9/9] [DON'T COMMIT] tests for all new arm cores Yann E. MORIN
` (2 subsequent siblings)
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
The armv8.1a generation is a cumulative extension to armv8a. It adds new
extensions, and makes some previously optional ones now mandatory.
Since gcc correctly enables the appropriate extensions based on the core
name, we don't really need to introduce a separate config for armv8.1a,
and we can piggyback on armv8a.
All those new cores are aarch64 only (gcc fails to build in arm mode).
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
arch/Config.in.arm | 35 +++++++++++++++++++++++++++++++++++
1 file changed, 35 insertions(+)
diff --git a/arch/Config.in.arm b/arch/Config.in.arm
index 40d6ae18b8..1cb75c1b69 100644
--- a/arch/Config.in.arm
+++ b/arch/Config.in.arm
@@ -400,6 +400,37 @@ config BR2_xgene1
select BR2_ARM_CPU_ARMV8A
select BR2_ARCH_HAS_MMU_OPTIONAL
select BR2_ARCH_NEEDS_GCC_AT_LEAST_5
+
+if BR2_ARCH_IS_64
+comment "armv8.1a cores"
+config BR2_thunderx2t99
+ bool "thunderx2t99"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_thunderx2t99p1
+ bool "thunderx2t99p1"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+config BR2_vulcan
+ bool "vulcan"
+ select BR2_ARM_CPU_HAS_ARM if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_NEON if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_THUMB2 if !BR2_ARCH_IS_64
+ select BR2_ARM_CPU_HAS_FP_ARMV8
+ select BR2_ARM_CPU_ARMV8A
+ select BR2_ARCH_HAS_MMU_OPTIONAL
+ select BR2_ARCH_NEEDS_GCC_AT_LEAST_7
+endif # BR2_ARCH_IS_64
endchoice
config BR2_ARM_ENABLE_NEON
@@ -705,6 +736,10 @@ config BR2_GCC_TARGET_CPU
default "thunderxt88" if BR2_thunderxt88
default "thunderxt88p1" if BR2_thunderxt88p1
default "xgene1" if BR2_xgene1
+ # armv8.1a
+ default "thunderx2t99" if BR2_thunderx2t99
+ default "thunderx2t99p1" if BR2_thunderx2t99p1
+ default "vulcan" if BR2_vulcan
config BR2_GCC_TARGET_ABI
default "aapcs-linux" if BR2_arm || BR2_armeb
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [PATCH 9/9] [DON'T COMMIT] tests for all new arm cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (7 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 8/9] arch/arm: add armv8.1a cores Yann E. MORIN
@ 2017-09-03 13:17 ` Yann E. MORIN
2017-09-03 13:24 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
2017-11-24 22:32 ` Thomas Petazzoni
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:17 UTC (permalink / raw)
To: buildroot
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
---
.../tests/toolchain/test_internal_aarch64_all.py | 162 +++++++++++++++++++++
.../tests/toolchain/test_internal_arm_all.py | 95 ++++++++++++
2 files changed, 257 insertions(+)
create mode 100644 support/testing/tests/toolchain/test_internal_aarch64_all.py
create mode 100644 support/testing/tests/toolchain/test_internal_arm_all.py
diff --git a/support/testing/tests/toolchain/test_internal_aarch64_all.py b/support/testing/tests/toolchain/test_internal_aarch64_all.py
new file mode 100644
index 0000000000..77f5ce8fdb
--- /dev/null
+++ b/support/testing/tests/toolchain/test_internal_aarch64_all.py
@@ -0,0 +1,162 @@
+import os
+import infra
+
+class InternalToolchainAarch64Base(infra.basetest.BRTest):
+ config = \
+ """
+ BR2_aarch64=y
+ BR2_INIT_NONE=y
+ BR2_SYSTEM_BIN_SH_NONE=y
+ # BR2_PACKAGE_BUSYBOX is not set
+ # BR2_TARGET_ROOTFS_TAR is not set
+ """
+
+################################################################################
+
+class TestInternalToolchainAarch64CortexA35(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_cortex_a35=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64CortexA73(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_cortex_a73=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64CortexA73A35(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_cortex_a73_a35=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64CortexA73A53(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_cortex_a73_a53=y
+ """
+
+ def test_run(self):
+ return True
+
+################################################################################
+
+class TestInternalToolchainAarch64Exynos1(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_exynos_m1=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64Falkor(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_falkor=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64Qdf24xx(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_qdf24xx=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderX(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderx=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderXt81(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderxt81=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderXt83(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderxt83=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderXt88(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderxt88=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderXt88p1(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderxt88p1=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64Xgene1(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_xgene1=y
+ """
+
+ def test_run(self):
+ return True
+
+################################################################################
+
+class TestInternalToolchainAarch64ThunderX2t99(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderx2t99=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64Vulcan(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_vulcan=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainAarch64ThunderX2t99p1(infra.basetest.BRTest):
+ config = InternalToolchainAarch64Base.config + \
+ """
+ BR2_thunderx2t99p1=y
+ """
+
+ def test_run(self):
+ return True
diff --git a/support/testing/tests/toolchain/test_internal_arm_all.py b/support/testing/tests/toolchain/test_internal_arm_all.py
new file mode 100644
index 0000000000..8c3811cc5d
--- /dev/null
+++ b/support/testing/tests/toolchain/test_internal_arm_all.py
@@ -0,0 +1,95 @@
+import os
+import infra
+
+class InternalToolchainArmBase(infra.basetest.BRTest):
+ config = \
+ """
+ BR2_arm=y
+ BR2_INIT_NONE=y
+ BR2_SYSTEM_BIN_SH_NONE=y
+ # BR2_PACKAGE_BUSYBOX is not set
+ # BR2_TARGET_ROOTFS_TAR is not set
+ """
+
+class TestInternalToolchainArmCortexA32(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_cortex_a32=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmCortexA35(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_cortex_a35=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmCortexA73(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_cortex_a73=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmCortexA73A35(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_cortex_a73_a35=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmCortexA73A53(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_cortex_a73_a53=y
+ """
+
+ def test_run(self):
+ return True
+
+################################################################################
+
+class TestInternalToolchainArmExynos1(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_exynos_m1=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmFalkor(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_falkor=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmQdf24xx(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_qdf24xx=y
+ """
+
+ def test_run(self):
+ return True
+
+class TestInternalToolchainArmXgene1(infra.basetest.BRTest):
+ config = InternalToolchainArmBase.config + \
+ """
+ BR2_xgene1=y
+ """
+
+ def test_run(self):
+ return True
--
2.11.0
^ permalink raw reply related [flat|nested] 51+ messages in thread* [Buildroot] [pull request] arch/arm: add some new armv8a cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (8 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [PATCH 9/9] [DON'T COMMIT] tests for all new arm cores Yann E. MORIN
@ 2017-09-03 13:24 ` Yann E. MORIN
2017-11-24 22:32 ` Thomas Petazzoni
10 siblings, 0 replies; 51+ messages in thread
From: Yann E. MORIN @ 2017-09-03 13:24 UTC (permalink / raw)
To: buildroot
All,
On 2017-09-03 15:17 +0200, Yann E. MORIN spake thusly:
> PS. I've sent this series as a real pull-request, just to see how
> the new patchwork handles it: currently, series are sorted by the
> first patch, and patchwork misses the cover-letter. Let's see if
> a cover-letter that is a pull-request gets noticed.
Definitely not working, and even breaks the series-detection of
patchowrk:
- it does not classifies the first patch in the series as being
part of the series;
- it treats the cover-letter as bieing on its own, totally
decorelated from the rest of the series.
Too bad...
Sorry for the noise... :-(
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'
^ permalink raw reply [flat|nested] 51+ messages in thread* [Buildroot] [pull request] arch/arm: add some new armv8a cores
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
` (9 preceding siblings ...)
2017-09-03 13:24 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
@ 2017-11-24 22:32 ` Thomas Petazzoni
10 siblings, 0 replies; 51+ messages in thread
From: Thomas Petazzoni @ 2017-11-24 22:32 UTC (permalink / raw)
To: buildroot
Hello,
On Sun, 3 Sep 2017 15:17:44 +0200, Yann E. MORIN wrote:
> Yann E. MORIN (9):
> arch/arm: re-order cores choice
> arch/arm: simplify hiding non 64-bit cores
> arch/arm: armv8 is really armv8a
Those patches applied to next.
> arch/arm: do not allow soft-float for armv8a
This one not applied, see my reply to it. I've kept it as "New" in
patchwork for now.
> arch/arm: add cortex-A32
> arch/arm: add some armv8a cortex variants
> arch/arm: add some non-cortex armv8a cores
> arch/arm: add armv8.1a cores
I've applied those four patches to next as well. As I replied to one of
them, I'm not super convinced exhaustively adding all cores is really
useful, but since you did it, and build tested all of them, I applied
the patches anyway.
Thanks!
Thomas Petazzoni
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
` (3 preceding siblings ...)
2017-09-03 13:17 ` [Buildroot] [pull request] arch/arm: add some new armv8a cores Yann E. MORIN
@ 2017-11-24 21:58 ` Thomas Petazzoni
4 siblings, 0 replies; 51+ messages in thread
From: Thomas Petazzoni @ 2017-11-24 21:58 UTC (permalink / raw)
To: buildroot
Hello,
On Sun, 3 Sep 2017 11:53:48 +0200, Yann E. MORIN wrote:
> Yann E. MORIN (3):
> arch/mips: inverse the NaN logic
> arch/mips: inverse the mfpxx logic
> toolchain/buildroot: glibc requires header >= 4.5 with NaN-2008
Series applied to next. Thanks!
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 51+ messages in thread
* [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version
2017-09-03 9:44 ` [Buildroot] [PATCH 0/8] arch: some require a minimal gcc version Yann E. MORIN
` (8 preceding siblings ...)
2017-09-03 9:53 ` [Buildroot] [PATCH 0/3] arch: fix MIPS NaN and floating-point handling Yann E. MORIN
@ 2017-11-24 21:23 ` Thomas Petazzoni
9 siblings, 0 replies; 51+ messages in thread
From: Thomas Petazzoni @ 2017-11-24 21:23 UTC (permalink / raw)
To: buildroot
Hello,
On Sun, 3 Sep 2017 11:44:28 +0200, Yann E. MORIN wrote:
> Yann E. MORIN (8):
> arch: introduce minimal required gcc version
> package/gcc: hide versions too old for the current arch
> toolchain/external-custom: hide versions too old for the current arch
> toolchain/external: hide versions too old for the current arch
> arch/bfin: needs gcc >= 6
> arch/mips: some variants need different gcc versions
> arch/arm: some variants need different gcc versions
> package/gcc: slight cleanup and reorg in remaining arch depends
Series applied to next, after fixing the minor conflicts that were
caused by other changes being merged in between, and fixing a few typos
in the commit log.
I'm still a bit worried about external toolchains that have gcc
versions containing backports that support for an architecture or
architecture variant that would normally require a more recent gcc
version.
For example, gcc support for ARC was added in gcc 7.x, so we will be
adding the statement that ARC needs gcc >= 7.x. However, the current
Synopsys external toolchain for ARC uses gcc 6.x, so it wouldn't be
selectable anymore.
I can imagine that Synopsys will rebuild their toolchain with gcc 7.x
soonish, so perhaps the problem will just disappear by itself.
But it might still be an issue in other situations. Let's see what
happens in the future! :-)
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 51+ messages in thread