Re: [PATCH 15/15] mmc: host: sdhci-iproc: Tell the compiler that ACPI functions may not be used

[Scott Branden]

On 2020-07-02 1:13 p.m., Lee Jones wrote:
On Thu, 02 Jul 2020, Scott Branden wrote:

Hi Lee,

On 2020-07-01 5:47 a.m., Lee Jones wrote:
... as is the case when !CONFIG_ACPI.

Fixes the following W=1 kernel build warning:

   mmc/host/sdhci-iproc.c:297:36: warning: ‘sdhci_iproc_acpi_ids’ defined but 
not used [-Wunused-const-variable=]

Cc: Adrian Hunter <adrian.hun...@intel.com>
Cc: Ray Jui <r...@broadcom.com>
Cc: Scott Branden <sbran...@broadcom.com>
Cc: bcm-kernel-feedback-l...@broadcom.com
Signed-off-by: Lee Jones <lee.jo...@linaro.org>
---
   drivers/mmc/host/sdhci-iproc.c | 2 +-
   1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mmc/host/sdhci-iproc.c b/drivers/mmc/host/sdhci-iproc.c
index 225603148d7de..9c7d130205edd 100644
--- a/drivers/mmc/host/sdhci-iproc.c
+++ b/drivers/mmc/host/sdhci-iproc.c
@@ -294,7 +294,7 @@ static const struct of_device_id sdhci_iproc_of_match[] = {
   };
   MODULE_DEVICE_TABLE(of, sdhci_iproc_of_match);
Would it better to compile this in if CONFIG_ACPI is defined instead?
#if CONFIG_ACPI
Literally can't win.

  
https://lore.kernel.org/linux-arm-kernel/CAGb2v66Ws4WNPZbOYQvikKoozj_2WjzS-Jq-o2VxT77=k0v...@mail.gmail.com/

Damned if we do, damned if we don't. ;)
This is very minor in this one change here, but it's just more bloat of 
things that really should be compiled out if they're not used.
I don't understand why __maybe_unused would be preferred approach - 
unless it is not linked in when it is not used?

-static const struct acpi_device_id sdhci_iproc_acpi_ids[] = {
+static const struct acpi_device_id __maybe_unused sdhci_iproc_acpi_ids[] = {
        { .id = "BRCM5871", .driver_data = (kernel_ulong_t)&iproc_cygnus_data },
        { .id = "BRCM5872", .driver_data = (kernel_ulong_t)&iproc_data },
        { /* sentinel */ }