| Message ID | 20210928195533.1736944-2-mmayer@broadcom.com |
|---|---|
| State | Accepted |
| Headers | show |
| Series | Build issue related to "command -v" | expand |
Reviewed-by: Petr Vorel <petr.vorel@gmail.com>
Kind regards,
Petr
On Tue, 28 Sep 2021 12:55:33 -0700 Markus Mayer via buildroot <buildroot@buildroot.org> wrote: > Set HOSTCC_NOCCACHE and HOSTCXX_NOCCACHE only if they are not set. This > allows recursive calls to "make" to work as intended in the presence of > ccache. > > Without guarding these variables, a recursive invocation of make would > re-define What is the use-case for a recursive invocation of make, reparsing the Buildroot Makefile? Thomas
On 28/12/2021 22:18, Thomas Petazzoni wrote: > On Tue, 28 Sep 2021 12:55:33 -0700 > Markus Mayer via buildroot <buildroot@buildroot.org> wrote: > >> Set HOSTCC_NOCCACHE and HOSTCXX_NOCCACHE only if they are not set. This >> allows recursive calls to "make" to work as intended in the presence of >> ccache. >> >> Without guarding these variables, a recursive invocation of make would >> re-define > > What is the use-case for a recursive invocation of make, reparsing the > Buildroot Makefile? The cover text mentions it: calling make legal-info inside a post-build script (or package, actually) to somehow include the result in the image.
On Tue, 28 Dec 2021 22:26:35 +0100 Nicolas Cavallari <nicolas.cavallari@green-communications.fr> wrote: > The cover text mentions it: calling make legal-info inside a post-build > script (or package, actually) to somehow include the result in the image. Many thanks for pointing it out, because the cover letter obviously contains all the details. I am looking at patches through patchwork, so I tend to only see the patches themselves and not necessarily an associated cover letter. Especially for single patches, where a cover letter is rarely present, and therefore I rarely tend to search for such a cover letter in the mailing list. Best regards, Thomas
Hello Markus, On Tue, 28 Sep 2021 12:55:33 -0700 Markus Mayer via buildroot <buildroot@buildroot.org> wrote: > Set HOSTCC_NOCCACHE and HOSTCXX_NOCCACHE only if they are not set. This > allows recursive calls to "make" to work as intended in the presence of > ccache. > > Without guarding these variables, a recursive invocation of make would > re-define > HOSTCC_NOCCACHE := $(HOSTCC) > and > HOSTCXX_NOCCACHE := $(HOSTCXX) > at a point in time when HOSTCC and HOSTCXX already point to ccache. > > Signed-off-by: Markus Mayer <mmayer@broadcom.com> > --- > Makefile | 4 ++++ > 1 file changed, 4 insertions(+) I have extended the commit log somewhat to give more details, and applied! Thanks a lot, Thomas
diff --git a/Makefile b/Makefile index d248bd76b2..71c0577030 100644 --- a/Makefile +++ b/Makefile @@ -286,12 +286,16 @@ ifndef HOSTCC HOSTCC := gcc HOSTCC := $(shell command -v $(HOSTCC) || type -p $(HOSTCC) || echo gcc) endif +ifndef HOSTCC_NOCCACHE HOSTCC_NOCCACHE := $(HOSTCC) +endif ifndef HOSTCXX HOSTCXX := g++ HOSTCXX := $(shell command -v $(HOSTCXX) || type -p $(HOSTCXX) || echo g++) endif +ifndef HOSTCXX_NOCCACHE HOSTCXX_NOCCACHE := $(HOSTCXX) +endif ifndef HOSTCPP HOSTCPP := cpp endif
Set HOSTCC_NOCCACHE and HOSTCXX_NOCCACHE only if they are not set. This allows recursive calls to "make" to work as intended in the presence of ccache. Without guarding these variables, a recursive invocation of make would re-define HOSTCC_NOCCACHE := $(HOSTCC) and HOSTCXX_NOCCACHE := $(HOSTCXX) at a point in time when HOSTCC and HOSTCXX already point to ccache. Signed-off-by: Markus Mayer <mmayer@broadcom.com> --- Makefile | 4 ++++ 1 file changed, 4 insertions(+)