Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors)
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks)
d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock)
2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate())
5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.)
8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.)
3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
40be0dd (net: add devm version of alloc_etherdev_mqs function)
34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it)
aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
36401cb (brcmfmac: be more verbose when PSM's watchdog fires)
9587a01 (brcmfmac: merge two brcmf_err macros into one)
087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors)
d063055 (brcmfmac: merge two remaining brcmf_err macros)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
91b6328 (brcmfmac: Use net_device_stats from struct net_device)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Cherry-picked and build tested on Linux 4.9.20 for ARCH=arm/arm64/i386/x86_64 + allmodconfig.
Regards, Amit Pundir
[1] https://github.com/lede-project/source. These patches are mostly cherry-picked from target/linux/*/patches-4.9 directories. Though some are also put in package/kernel/*/patches/ directory.
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
thanks,
greg k-h
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well? Lede cherry-picked few of these on 4.4 too, but I'll look into 4.4 story of lede patches later.
Patches in this list are to be applied from top to bottom.
As discussed in my earlier email, I'm sending you another list of somewhat random but seemingly useful upstream commits list shortly. That list of patches is being build tested on top of these lede patches.
Regards, Amit Pundir
thanks,
greg k-h
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
IIRC, these are to be sent to stable@ as a patch series. See Sumit's email as an example -
https://www.spinics.net/lists/stable/msg164490.html
- ssp
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
Lede cherry-picked few of these on 4.4 too, but I'll look into 4.4 story of lede patches later.
That's good, thanks,
greg k-h
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Missed it by a minute or two. I have just sent the patches on stable https://www.spinics.net/lists/stable/msg165892.html. I'll send the git commit ids from next time, unless I'm updating the original patch.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
I'll prepare the list and send it across.
Regards, Amit Pundir
Lede cherry-picked few of these on 4.4 too, but I'll look into 4.4 story of lede patches later.
That's good, thanks,
greg k-h
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
For 4.10, I'll send the list of patches based on the lede stable changes which make it to stable-queue for 4.9. Because few patches for-4.9 are NACKed by Rafał Miłecki already. Quoting him here, "Some of these patches are clean ups, not a really important fixes." Though he didn't mention which patches he meant. I checked back the list of patches again and assume he meant following patches of his and others:
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors) The only reason I'm keeping this ^^ cleanup patch in the list is because it helps following patch (8d2cc5cc) apply cleanly.
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Build tested on arm/arm64 + allmodconfig.
Regards, Amit Pundir
Lede cherry-picked few of these on 4.4 too, but I'll look into 4.4 story of lede patches later.
That's good, thanks,
greg k-h
On Wed, Apr 05, 2017 at 01:55:59PM +0530, Amit Pundir wrote:
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
For 4.10, I'll send the list of patches based on the lede stable changes which make it to stable-queue for 4.9. Because few patches for-4.9 are NACKed by Rafał Miłecki already. Quoting him here, "Some of these patches are clean ups, not a really important fixes." Though he didn't mention which patches he meant. I checked back the list of patches again and assume he meant following patches of his and others:
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
Why not ask on the list to be sure?
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors) The only reason I'm keeping this ^^ cleanup patch in the list is because it helps following patch (8d2cc5cc) apply cleanly.
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Build tested on arm/arm64 + allmodconfig.
Again, please post on stable@ so that others can take advantage of this, and review them.
thanks,
greg k-h
On 5 April 2017 at 14:18, Greg KH gregkh@google.com wrote:
On Wed, Apr 05, 2017 at 01:55:59PM +0530, Amit Pundir wrote:
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote:
Hi Greg,
For your consideration following upstream patches taken from lede source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
For 4.10, I'll send the list of patches based on the lede stable changes which make it to stable-queue for 4.9. Because few patches for-4.9 are NACKed by Rafał Miłecki already. Quoting him here, "Some of these patches are clean ups, not a really important fixes." Though he didn't mention which patches he meant. I checked back the list of patches again and assume he meant following patches of his and others:
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
Why not ask on the list to be sure?
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors) The only reason I'm keeping this ^^ cleanup patch in the list is because it helps following patch (8d2cc5cc) apply cleanly.
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Build tested on arm/arm64 + allmodconfig.
Again, please post on stable@ so that others can take advantage of this, and review them.
I was preparing v2 of my proposed lede stable-4.9 commits to stable and then I see your email just now on stable@vger. Have you started pulling the patches already? Should I not send v2 now? Please confirm.
Regards, Amit Pundir
thanks,
greg k-h
On 5 April 2017 at 15:48, Amit Pundir amit.pundir@linaro.org wrote:
On 5 April 2017 at 14:18, Greg KH gregkh@google.com wrote:
On Wed, Apr 05, 2017 at 01:55:59PM +0530, Amit Pundir wrote:
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote: > Hi Greg, > > For your consideration following upstream patches taken from lede > source tree[1] targeted for 4.9.y.
<snip>
Can you send this to the stable@ list instead so that everyone can see it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
Also, are any of these to be applied to 4.10? And, what order do these apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
For 4.10, I'll send the list of patches based on the lede stable changes which make it to stable-queue for 4.9. Because few patches for-4.9 are NACKed by Rafał Miłecki already. Quoting him here, "Some of these patches are clean ups, not a really important fixes." Though he didn't mention which patches he meant. I checked back the list of patches again and assume he meant following patches of his and others:
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
Why not ask on the list to be sure?
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors) The only reason I'm keeping this ^^ cleanup patch in the list is because it helps following patch (8d2cc5cc) apply cleanly.
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Build tested on arm/arm64 + allmodconfig.
Again, please post on stable@ so that others can take advantage of this, and review them.
I was preparing v2 of my proposed lede stable-4.9 commits to stable and then I see your email just now on stable@vger. Have you started pulling the patches already? Should I not send v2 now? Please confirm.
Sent v2 anyway. Thanks.
Regards, Amit Pundir
thanks,
greg k-h
On Wed, Apr 05, 2017 at 04:04:49PM +0530, Amit Pundir wrote:
On 5 April 2017 at 15:48, Amit Pundir amit.pundir@linaro.org wrote:
On 5 April 2017 at 14:18, Greg KH gregkh@google.com wrote:
On Wed, Apr 05, 2017 at 01:55:59PM +0530, Amit Pundir wrote:
On 4 April 2017 at 11:46, Greg KH gregkh@google.com wrote:
On Mon, Apr 03, 2017 at 11:57:22PM +0530, Amit Pundir wrote:
On 3 April 2017 at 23:31, Greg KH gregkh@google.com wrote: > On Mon, Apr 03, 2017 at 11:06:23PM +0530, Amit Pundir wrote: >> Hi Greg, >> >> For your consideration following upstream patches taken from lede >> source tree[1] targeted for 4.9.y. > > <snip> > > Can you send this to the stable@ list instead so that everyone can see > it there?
Sure. Should I send them in the same format as this one i.e. only git commit ID list?
Either is fine, a list of git commit ids is simple to handle for me.
> Also, are any of these to be applied to 4.10? And, what order do these > apply in, the list you gave, or backwards?
I did not apply them on 4.10 as such but I'm sure few of them are applicable to 4.10. Should I send you a patch list for 4.10 as well?
Yes please.
For 4.10, I'll send the list of patches based on the lede stable changes which make it to stable-queue for 4.9. Because few patches for-4.9 are NACKed by Rafał Miłecki already. Quoting him here, "Some of these patches are clean ups, not a really important fixes." Though he didn't mention which patches he meant. I checked back the list of patches again and assume he meant following patches of his and others:
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
Why not ask on the list to be sure?
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
09f3510 (ARM: BCM5301X: Add back handler ignoring external)
0c2bf9f (ARM: dts: BCM5301X: Correct GIC_PPI interrupt flags)
6e347b5 (PCI: iproc: Save host bridge window resource in struct)
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
e247454 (i2c: bcm2835: Fix hang for writing messages larger than 16 bytes)
d4030d7 (i2c: bcm2835: Protect against unexpected TXW/RXR interrupts)
23c9540 (i2c: bcm2835: Use dev_dbg logging on transfer errors) The only reason I'm keeping this ^^ cleanup patch in the list is because it helps following patch (8d2cc5cc) apply cleanly.
8d2cc5cc (i2c: bcm2835: Can't support I2C_M_IGNORE_NAK)
2201ac6 (dmaengine: bcm2835: Fix cyclic DMA period splitting)
cd4b1e3 (usb: dwc2: Remove unnecessary kfree)
bd5d213 (mtd: bcm47xxpart: fix parsing first block after aligned TRX)
3ec7544 (of: Add check to of_scan_flat_dt() before accessing initial_boot_params)
7272416 (rt2500usb: don't mark register accesses as inline)
f4737a6 (brcmfmac: check brcmf_bus_get_memdump result for error)
93c7018 (rt2x00usb: do not anchor rx and tx urb's)
0488a61 (rt2x00usb: fix anchor initialization)
a083c8f (rt2x00: Fix incorrect usage of CONFIG_RT2X00_LIB_USB)
6232c17 (rt2x00: avoid introducing a USB dependency in the rt2x00lib module)
Build tested on arm/arm64 + allmodconfig.
Again, please post on stable@ so that others can take advantage of this, and review them.
I was preparing v2 of my proposed lede stable-4.9 commits to stable and then I see your email just now on stable@vger. Have you started pulling the patches already? Should I not send v2 now? Please confirm.
Sent v2 anyway. Thanks.
Not a problem, thanks for sending them, I'll review them in the next few days...
greg k-h
On 5 April 2017 at 13:55, Amit Pundir amit.pundir@linaro.org wrote: <snip>
5d1f2d2 (ARM: dts: BCM5301X: Set 5 GHz wireless frequency limits) 155e8b3 (clk: bcm: Support rate change propagation on bcm2835 clocks) d86d46a (clk: bcm: Allow rate change propagation to PLLH_AUX on VEC clock) 2aab7a2 (clk: bcm: Fix 'maybe-uninitialized' warning in bcm2835_clock_choose_div_and_prate()) 5548609 (clk: bcm2835: Don't rate change PLLs on behalf of DSI PLL dividers.) 8a39e9f (clk: bcm2835: Register the DSI0/DSI1 pixel clocks.) 3f91958 (clk: bcm2835: Add leaf clock measurement support, disabled by default) 40be0dd (net: add devm version of alloc_etherdev_mqs function) 34a5102 (net: bgmac: allocate struct bgmac just once & don't copy it) aa8863e (net: bgmac: drop struct bcma_mdio we don't need anymore) 36401cb (brcmfmac: be more verbose when PSM's watchdog fires) 9587a01 (brcmfmac: merge two brcmf_err macros into one) 087fa71 (brcmfmac: switch to C function (__brcmf_err) for printing errors) d063055 (brcmfmac: merge two remaining brcmf_err macros) 91b6328 (brcmfmac: Use net_device_stats from struct net_device)
So I dropped above set of patches and below is the new list of patches from lede for stable-4.9 consideration. To be cherry-picked from top to bottom:
Turned out all the patches which got rejected for stable today were already filtered out above and I shouldn't have sent them in v2. I'll be extra vigilant from next time.
<snip>
6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
What is the Revert policy on stable? This above patch was NACKed by one of the reviewers but made it through to 4.4, 4.9 and 4.10 stable. There was some discussion around this patch but we should have been more explicit about it and should have informed you earlier. This patch is meant for a MIPS IRQ stack bug introduced in 4.11. So it is not applicable to 4.4, 4.9 and 4.10 stable.
Regards, Amit Pundir
On Wed, Apr 12, 2017 at 11:30:55PM +0530, Amit Pundir wrote:
<snip> > 6c356ed (MIPS: Lantiq: Fix cascaded IRQ setup)
What is the Revert policy on stable? This above patch was NACKed by one of the reviewers but made it through to 4.4, 4.9 and 4.10 stable. There was some discussion around this patch but we should have been more explicit about it and should have informed you earlier. This patch is meant for a MIPS IRQ stack bug introduced in 4.11. So it is not applicable to 4.4, 4.9 and 4.10 stable.
Tell me, on the stable list, and I will revert the commit in the next release. Ideally if you can provide the git ids that I need to revert, that would save me having to look them up :)
thanks,
greg k-h
-
Amit Pundir -
Greg KH -
Sandeep Patil