November 2014 - kernelnewbies@kernelnewbies.org

Healthy cpu operating temperature range
by Puneet Agarwal Nov. 4, 2014

Nov. 4, 2014

Hello, I am having some servers which serves http traffic at high qps. Strangely for past few days I am seeing latency going up. I am worried if these CPU temperatures can be the reason. Core 0: +74.0°C (high = +90.0°C, crit = +100.0°C) Core 1: +78.0°C (high = +90.0°C, crit = +100.0°C) Core 9: +73.0°C (high = +90.0°C, crit = +100.0°C) Core 10: +81.0°C (high = +90.0°C, crit = +100.0°C) coretemp-isa-0004 Adapter: ISA adapter Core 0: +81.0°C (high = +90.0°C, crit = +100.0°C) Core 1: +80.0°C (high = +90.0°C, crit = +100.0°C) Core 9: +81.0°C (high = +90.0°C, crit = +100.0°C) Core 10: +81.0°C (high = +90.0°C, crit = +100.0°C) top command gives load average, Load: 12.03, 6.76, 4.44 Processor specification sheet says Tcase is 63.1°C, and the sensor tool states 90°C as high. This is not letting me believe which one is correct to pick and what is the ideal temperature range for CPU to run. Thanks and Regards

1 0

[PATCH] staging: Fix spacing between function name and parentheses
by Nicholas Krause Nov. 4, 2014

Nov. 4, 2014

Fixes checkpatch coding style warning about unneeded space between function name an parentheses. Signed-off-by: Nicholas Krause <xerofoify(a)gmail.com> --- Untested drivers/staging/octeon-usb/octeon-hcd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/staging/octeon-usb/octeon-hcd.c b/drivers/staging/octeon-usb/octeon-hcd.c index 5f9db4c..bbeb0cc 100644 --- a/drivers/staging/octeon-usb/octeon-hcd.c +++ b/drivers/staging/octeon-usb/octeon-hcd.c @@ -412,7 +412,7 @@ struct octeon_hcd { type c; \ while (1) { \ c.u32 = __cvmx_usb_read_csr32(usb, address); \ - if (c.s.field op (value)) { \ + if (c.s.field op(value)) { \ result = 0; \ break; \ } else if (cvmx_get_cycle() > done) { \ -- 1.9.1

15 25

build a kernel over a mounted NFS partition
by Kevin Wilson Nov. 3, 2014

Nov. 3, 2014

Hi, I have a question regarding building a kernel over a mounted NFS partition. I can arrange such a setup, where an NFS server with 1Gb nic will be connected to a very strong server (but where I do not have enough stoage for putting a kernel + it object files). The question is - does it worth it ? or building by NFS is too much expensive because of the overhead ? more details: The strong server, on which I intend to build, has about 4 time more CPU power than the NFS server. My question is: what is the expected behaviour ? is the fact that I am building on that strong server over a 1GB network connection incurs such a performance cost so that I will have about two times (over even more, like 3 times or more) than building the kernel on the strong machine ? Regards, Kevin

2 1

[PATCH] staging: bcm: vendorspecificextn: Fix coding style
by Sunil Shahu Nov. 3, 2014

Nov. 3, 2014

Hello, I was going through kernel source and found out that I can work on staging area for fixes. I have generated a patch for coding style fixes. I have few questions about how do submit my patch and to whom. 1) Is it ok if I send it directly to Greg Kroah-Hartman <gregkh(a)linuxfoundation.org>, who seems to be signing off all the patches in that directory as per git log. 2) What is the standard procedure to submit my patch after I have verified it. Whom do I send it? How do I get it review? How to get it merged in mainline? Also, I am attaching my patch here. Any suggestions on patch and answer to my queries will be great help. Thanks, Sunil.

3 2

copy_from_user user space address mapping in the kernel page table.
by mind entropy Nov. 3, 2014

Nov. 3, 2014

Hi, In the copy_from_user when the user passes the virtual address is the address mapped in the kernel page tables? In the function there is a source is a user virtual address(<3GB) and the destination is the kernel virtual address(3GB - 4GB) and for the MMU to access the physical page frame when in kernel space operating in behalf of the process context there should be a mapping of the virtual address to the page frame, correct? I am assuming the implementation would be to take the virtual address and walk the page table to get to the physical page frame and map the physical page frame in the kernel page table (i.e. swapper_pgd). Is this correct? If not what would be the mechanism? Thanks, Gautam.

2 2

little at eudyptula-challenge.org
by Manohar Singh Nov. 3, 2014

Nov. 3, 2014

-- Regards, Manohar Singh Aswal. +91-9916523441

1 0

Fwd: Re: [PATCH] staging: Fix checkpatch error messages about unneeded space before commas in ms.c
by nick Nov. 2, 2014

Nov. 2, 2014

-------- Forwarded Message -------- Subject: Re: [PATCH] staging: Fix checkpatch error messages about unneeded space before commas in ms.c Date: Sun, 2 Nov 2014 15:56:36 +0100 From: Peter Senna Tschudin <peter.senna(a)gmail.com> To: Nicholas Krause <xerofoify(a)gmail.com> This is good! You can try to submit it. On Tue, Oct 28, 2014 at 9:25 PM, Nicholas Krause <xerofoify(a)gmail.com> wrote: > Fixes various checkpatch error messages about removing unneeded spaces before commas on > certain lines in ms.c. > > Signed-off-by: Nicholas Krause <xerofoify(a)gmail.com> > --- > drivers/staging/rts5208/ms.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/drivers/staging/rts5208/ms.c b/drivers/staging/rts5208/ms.c > index 228e483..d77153c 100644 > --- a/drivers/staging/rts5208/ms.c > +++ b/drivers/staging/rts5208/ms.c > @@ -781,7 +781,7 @@ static int msxc_change_power(struct rtsx_chip *chip, u8 mode) > buf[4] = 0; > buf[5] = 0; > > - retval = ms_write_bytes(chip, PRO_WRITE_REG , 6, NO_WAIT_INT, buf, 6); > + retval = ms_write_bytes(chip, PRO_WRITE_REG, 6, NO_WAIT_INT, buf, 6); > if (retval != STATUS_SUCCESS) > TRACE_RET(chip, STATUS_FAIL); > > @@ -1291,7 +1291,7 @@ static int ms_write_extra_data(struct rtsx_chip *chip, > for (i = 6; i < MS_EXTRA_SIZE + 6; i++) > data[i] = buf[i - 6]; > > - retval = ms_write_bytes(chip, WRITE_REG , (6+MS_EXTRA_SIZE), > + retval = ms_write_bytes(chip, WRITE_REG, (6+MS_EXTRA_SIZE), > NO_WAIT_INT, data, 16); > if (retval != STATUS_SUCCESS) > TRACE_RET(chip, STATUS_FAIL); > @@ -1342,7 +1342,7 @@ static int ms_read_page(struct rtsx_chip *chip, u16 block_addr, u8 page_num) > data[4] = 0x20; > data[5] = page_num; > > - retval = ms_write_bytes(chip, WRITE_REG , 6, NO_WAIT_INT, data, 6); > + retval = ms_write_bytes(chip, WRITE_REG, 6, NO_WAIT_INT, data, 6); > if (retval != STATUS_SUCCESS) > TRACE_RET(chip, STATUS_FAIL); > > @@ -1619,7 +1619,7 @@ static int ms_copy_page(struct rtsx_chip *chip, u16 old_blk, u16 new_blk, > data[4] = 0x20; > data[5] = i; > > - retval = ms_write_bytes(chip, WRITE_REG , 6, NO_WAIT_INT, > + retval = ms_write_bytes(chip, WRITE_REG, 6, NO_WAIT_INT, > data, 6); > if (retval != STATUS_SUCCESS) > TRACE_RET(chip, STATUS_FAIL); > @@ -1988,7 +1988,7 @@ RE_SEARCH: > RTSX_WRITE_REG(chip, PPBUF_BASE2, 0xFF, 0x88); > RTSX_WRITE_REG(chip, PPBUF_BASE2 + 1, 0xFF, 0); > > - retval = ms_transfer_tpc(chip, MS_TM_WRITE_BYTES, WRITE_REG , 1, > + retval = ms_transfer_tpc(chip, MS_TM_WRITE_BYTES, WRITE_REG, 1, > NO_WAIT_INT); > if (retval != STATUS_SUCCESS) > TRACE_RET(chip, STATUS_FAIL); > -- > 1.9.1 > -- Peter

1 0

On spin_lock_irq_save and disabling specific interrupts.
by mind entropy Nov. 2, 2014

Nov. 2, 2014

Hi, When I do spin_lock_irq it disables IRQ's which were enabled. What is the reason for disabling all the irq's rather than disabling the specific IRQ line which shares the resource? Wouldn't this be more efficient and would not lose any interrupts especially if I have a critical sensor ADC? Thanks, Gautam.

2 1