drivers/pci: stop using LOG_DBG()

When the console UART is a PCI device, and PCI debug logging is enabled,
the system crashes because the UART is initialized before logging, but
the UART initialization invokes the PCI subsystem which invokes logging.
Reordering the initialization sequence will not fix this chicken/egg.

Luckily, the LOG_DBG() calls in the PCI subsystem appear to be bitrot
leftovers from early development, so they are simply removed.

Also mark myself as the owner of the PCI subsystem.

Fixes: #14763

Signed-off-by: Charles E. Youse <charles.youse@intel.com>
This commit is contained in:
Charles E. Youse 2019-03-26 10:51:42 -07:00 committed by Carles Cufí
parent 20e4ca48c7
commit eaf1c3b8a6
2 changed files with 1 additions and 8 deletions

View file

@ -118,6 +118,7 @@
/drivers/led/ @Mani-Sadhasivam
/drivers/led_strip/ @mbolivar
/drivers/modem/ @mike-scott
/drivers/pci/ @gnuless
/drivers/pinmux/stm32/ @rsalveti @idlethread
/drivers/sensor/ @bogdan-davidoaia @MaureenHelm
/drivers/sensor/hts*/ @avisconti

View file

@ -79,10 +79,6 @@
#include <soc.h>
#define LOG_LEVEL CONFIG_PCI_LOG_LEVEL
#include <logging/log.h>
LOG_MODULE_REGISTER(pci);
#include <pci/pci_mgr.h>
#include <pci/pci.h>
@ -260,7 +256,6 @@ static int pci_dev_scan(union pci_addr_reg pci_ctrl_addr,
/* Skip single function device */
if (lookup.func != 0 && !multi_function) {
LOG_DBG("Skip single function device");
break;
}
@ -399,7 +394,6 @@ int pci_bus_scan(struct pci_dev_info *dev_info)
for (; lookup.bus < lookup.buses; lookup.bus++) {
for (; lookup.dev < LSPCI_MAX_DEV; lookup.dev++) {
if (lookup.bus == 0 && lookup.dev == 0) {
LOG_DBG("Skip Host Bridge");
continue;
}
@ -434,8 +428,6 @@ static void pci_set_command_bits(struct pci_dev_info *dev_info, u32_t bits)
pci_ctrl_addr.field.device = dev_info->dev;
pci_ctrl_addr.field.reg = 1;
LOG_DBG("bits 0x%x", bits);
pci_read(DEFAULT_PCI_CONTROLLER, pci_ctrl_addr,
sizeof(u16_t), &pci_data);